1
INVESTIGATIONS OF OXYGEN REDUCTION REACTIONS IN NON-
AQUEOUS ELECTROLYTES AND THE LITHIUM-AIR BATTERY
A Dissertation Presented
by
Cormac Míchéal Ó’Laoire
to
The Department of Chemistry and Chemical Biology
in partial fulfillment of the requirements
for the Degree of
Doctor of Philosophy
in the field of
Chemistry
Northeastern University
Boston, Massachusetts
April, 2010
2
© Cormac M. Ó’Laoire
2010
All Rights Reserved
3
INVESTIGATIONS OF OXYGEN REDUCTION REACTIONS IN NON-
AQUEOUS ELECTROLYTES AND THE LITHIUM-AIR BATTERY
by
Cormac Míchéal Ó’Laoire
ABSTRACT OF DISSERTATION
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemistry
in the Graduate School of Arts and Sciences of Northeastern University, April 2010
4
Abstract
Unlocking the true energy capabilities of the lithium metal negative electrode
in a lithium battery has until now been limited by the low capacity intercalation and
conversion reactions at the positive electrodes. This is overcome by removing these
electrodes and allowing lithium to react directly with oxygen in the atmosphere
forming the Li-air battery. The Li/O2 battery redox couple has a theoretical specific
energy of 5200Wh/Kg and represents the ultimate energy density, environmentally
friendly battery.
Chapter 2 discusses the intimate role of electrolyte, in particular the role of ion
conducting salts on the mechanism and kinetics of oxygen reduction in non-aqueous
electrolytes designed for such applications and in determining the reversibility of the
electrode reactions. Such fundamental understanding of this high energy density
battery is crucial to harnessing its full energy potential. The kinetics and mechanisms
of O2 reduction in solutions of hexafluorophosphate salts of the general formula X+
PF6-, where, X = tetra butyl ammonium (TBA), K, Na and Li, in acetonitrile have
been studied on glassy carbon electrodes using cyclic voltammetry (CV) and rotating
disk electrode (RDE) techniques. Our results show that cation choice strongly
influences the reduction mechanism of O2. Large cations such as TBA facilitate
reversible O2 reduction involving the one electron reduction product, O2- which is
stabilized by the large TBA cation.. In contrast small cations like Li (and other alkali
metals), promote an irreversible electrochemical reaction. The initial reaction again
is one-electron reduction of O2 to LiO2 or other alkali metal superoxides. The LiO2
5
formed initially either decomposes to Li2O2 or undergoes further reduction to Li2O2
and Li2O. Electrochemical data supports the view that alkali metal oxides formed via
electrochemical and chemical reactions passivate the electrode surface inhibiting the
kinetics and reversibility of the processes. The O2 reduction mechanisms in the
presence of the different cations have been supplemented by kinetic parameters
determined from detailed analyses of the CV and RDE data. The Lewis acid
characteristics of the cation appear to be crucial in determining the reversibility of the
system. The organic solvent present in the Li+-conducting electrolyte has a major
role on the reversibility of each of the O2 reduction products as found from the work
discussed in the next chapter.
A fundamental study of the influence of solvents on the oxygen reduction
reaction (ORR) in a variety of non-aqueous electrolytes was conducted in chapter 4.
In this work special attention was paid to elucidate the mechanism of the oxygen
electrode processes in the rechargeable Li-air battery. Towards this end, using either
tetrabutylammonium hexafluorophosphate (TBAPF6) or lithium hexafluorophosphate
(LiPF6) electrolyte solutions in four different solvents, namely, dimethyl sulfoxide
(DMSO), acetonitrile (MeCN), dimethoxyethane (DME), and tetraethylene glycol
dimethyl ether (TEGDME), possessing a range of properties, we have determined that
the solvent and the supporting electrolyte cations in the solution act in concert to
influence the nature of reduction products and their rechargeability. In solutions
containing TBA+, O2 reduction is a highly reversible one-electron process involving
the O2/O2- couple in all of the electrolytes examined with little effect on the nature of
the solvent. On the other hand, in Li+-containing electrolytes relevant to the Li-air
6
battery, O2 reduction proceeds in a stepwise fashion to form O2-, O2
2- and O2- as
products. These reactions in presence of Li+ are irreversible or quasi-reversible
electrochemical processes and the solvents have significant influence on the kinetics,
and reversibility or lack thereof, of the different reduction products. The stabilization
of the one-electron reduction product, superoxide (O2-) in TBA+ solutions in all of the
solvents examined can be explained using Pearson’s Hard Soft Acid Base (HSAB)
theory involving the formation of the TBA+---O2- complex. The HSAB theory
coupled with the relative stabilities of the Li+-(solvent)n complexes existing in the
different solvents also provide an explanation for the different O2 reduction products
formed in Li+-conducting electrolyte solutions. Reversible reduction of O2 to long-
lived superoxide in a Li+-conducting electrolyte in DMSO has been shown for the
first time here.
Chapter 5 is the culmination of the thesis where the practical application of
the work is demonstrated.. We designed electrolytes that facilitate Li-Air
rechargeability, by applying the knowledge gained from chapters 2-4. A rechargeable
Li-air cell utilizing an electrolyte composed of a solution of LiPF6 in tetraethylene
glycol dimethyl ether, CH3O(CH2CH2O)4CH3 was designed, built and its performance
studied. It was shown that the cell yields high capacity and can be recharged in spite
the absence of catalyst in the carbon cathode. From the X-ray diffraction patterns of
the discharged carbon electrodes, the discharge product of the cell was identified to
be Li2O2 during normal discharge to 1.5 V. Discharging the cell to 1.0 V produces
Li 2O as well. The application of X-ray diffraction to identify these products formed
in a porous carbon electrode is shown here for the first time. The rechargeability of
7
the cell was investigated by repeated charge/discharge cycling of the cell, and the
factors limiting the cycle life of the cell were studied using AC impedance spectra of
the cells as a function of cycle number.
In conclusion, the work carried out in this research has shown that the O2
electrochemistry in organic electrolytes is substantially different from that in aqueous
electrolytes. Our work has uncovered the key roles the ion conducting salts and the
organic solvents play in determining the nature of the reduction products and their
reversibility. The results presented here for the first time provide a rational approach
to the design and selection of organic electrolyte solutions for use in the rechargeable
Li-air battery. Factors affecting the cycle life limitations of the Li-air cell have been
identified from the cycling performance and the associated impedance changes of
Li/air laboratory test cells. Our work is expected to contribute to the rapid
development of the rechargeable Li-air battery.
8
Acknowledgements
I would like to thank Dr K.M Abraham, for his mentorship and guidance during the
last six years of my graduate career. It is a great honor to be his first graduate student
and the first of what I am sure will be a long line of students. I’m grateful for K.M
taking me under his wing and selflessly imparting decades of scientific and life
advice. His uncanny knack of turning lemons into lemonade is inspiration to all his
students and serves as a life lesson to us all.
I thank Professor Sanjeev Mukerjee for his selflessness, patience and most of
all his uncompromising belief in his students. No question was too silly for Sanjeev
and no time was too bad to ask for help. His wonderful group is more like a family to
us foreign born students. Thanks for keeping our belly’s full and minds busy. I
appreciate all of the opportunities my advisors have given me, not only the freedom
in my research pursuits, but also the travel and networking experiences.
I would like to thank my thesis committee, Prof. David Budil, Prof. Max
Diem, Prof. Eugene Smotkin and Prof Graham Jones for their time and helpful
suggestions. I would also like to thank my Chemistry department who include but are
not limited to Nancy Weston, Jean Harris, Rich Pumphrey and Ed Witten.
The ever expanding list of past and present Mukerjee group members are too
numerous to list here, I give all of you my heartfelt gratitude without you I would not
be here. I thank Dr Tom Arruda and Dr Jamie Lawton embarking upon this journey
and seeing me through to the finish line. I thank Dr Joe Ziegelbauer for catalyzing
my research career. Thank you Matt, Chris, Naggapan, Cara, Brian, WenWen and
Vivek. Late night conversations with Dr Nazih Hakim were always enjoyable.
9
To my family the mighty expanse of the Atlantic could do little to dim the
bright glow of your love and support. Time has tested our resolve; I hope now all the
sacrifice has finally paid off. I thank my parents for making sure I always got home
during the summer and Christmas holidays. We have had some wonderful family
trips around New England and I look forward to many more. To Donncha my best
friend, I left you when you were a boy now I return you are a man. I wish you the
very best in your endeavors and look forward to you eclipsing my feats.
Airt, Lyndsey and little baby Sunneva you cannot underestimate how your
messages of support and calls warm my heart. Airt you really play the older big
brother role to perfection.. You are an inspiration to myself and Don, I know I can
always call on you for help or advice. When I hear “Whiskey in the jar” I always
think of you, the line “If Anyone can aid me it’s my brother in the army” really nails
our bond.
My dearest sister Nessa, I could not be prouder, your success inspires me.
One of the reasons I’m graduating is to make sure I get out before you, what would
mom and dad think if you got out first. I really appreciate your regular messages
letting know everything that’s happening at home. David I’m looking forward to
wedding and free dive lessons once you join the family.
To my wonderful parents Denis and Eucharia you can start cashing in all
those IOU’s. Hopefully it goes without saying how much I appreciate all your love.
Thank you for enabling me to pursue my dream and supporting me during the highs
and lows. I want you to know this is your PhD too.
10
To my breathtakingly beautiful Liz, anything I write could not express how I
much I love and value you. Thank you for your unwavering belief in me when I
doubted so many times. I was a long term project when you took me on, I thank you
for making me a better man.
Unfortunately not everyone can see me graduate, but I know they are with me
in spirit. I thank my grandparents Paddy & Monica O’Leary for all their love and
encouragement. I thank Dr Eugene Cheng who took me in and made me feel part of
his family. For all the great times we had, numerous banquets and massway band
concerts. Thank you for sharing your love of music and guitars with me. Walking
around Chinatown with you opened up so many doors for us
For the family member funerals I missed Damien O Callaghan and Ben Healy
may you all rest in peace.
11
Table of Contents
Abstract 4
Acknowledgements 8
Table of Contents 11
List of Tables 14
List of Figures 16
List of Illustrations 22
List of Abbreviations and Symbols 23
Chapter 1 Introduction 25
1.1 Energy Challenge 25
1.2 Batteries 27
1.3 Fundamentals of the Lithium –Air Battery 34
1.4 Lithium-Air Today 37
1.5 Non-Aqueous Electrolytes 40
1.6 Non Aqueous Oxygen Reduction Reaction (ORR) 45
1.7 Scope of Dissertation 47
1.8 References 48
Chapter 2 Electrochemical Studies of Ferrocene in a Lithium Ion
Conducting Organic Carbonate Electrolyte 51
2.1 Introduction 51
2.2 Experimental 53
2.2.1 Chemical Reagents 53
2.2.2 Instrumentation 54
2.3 Results and Discussion 55
2.3.1 Cyclic Voltammetry 55
2.3.2 Rotating Disk Electrode 62
2.4 Conclusions 66
12
2.5 References 68
Chapter 3 Elucidating the Mechanism of Oxygen Reduction for
Lithium-Air Battery Application 70
3.1 Introduction 70
3.2 Experimental 73
3.2.1 Chemical Reagents 73
3.2.2 Instrumentation 73
3.3 Results and Discussion 74
3.3.1 Oxygen Reduction in Tetrabutylammonium
hexafluorophosphate (TBA+PF-6)-
Based Electrolytes 76
3.3.2 Oxygen Reduction in Alkali Metal-
Hexafluorophosphate (X+PF-6)
Based Electrolytes 87
3.4 Conclusions 99
3.5 References 100
Chapter 4 Influence of Non-aqueous Solvents on the Electrochemistry
of Oxygen in the Rechargeable Lithium-Air battery 102
4.1 Introduction 102
4.2 Experimental 106
4.2.1 Materials 106
4.2.2 Electrochemical Experiments 106
4.3 Results and Discussion 108
4.3.1 ORR in Selected Non-Aqueous Electrolytes. 108
4.3.2 ORR in TBAPF6 solutions in DMSO, DME
and MeCN 110
4.3.3 ORR in LiPF6 solutions in DMSO, DME,
MeCN, and TEGDME 117
13
4.3.4 Impedance spectroscopy to determine O2
reduction kinetics 128
4.3.5 Understanding ORR in non-aqueous
electrolytes using Pearson’s HSAB Theory
4.4 Conclusions 136
4.5 References 137
Chapter 5 A Rechargeable Lithium/TEGDME-LiPF6/O2 Battery 140
5.1 Introduction 141
5.2 Experimental 141
5.2.1 Materials 141
5.2.2 Li/O2 Cells 142
5.3 Results and Discussion 143
5.3.1 Li/O2 Cell Discharge and Charge Behavior 146
5.3.2 Factors affecting the Cycle Life of the Li/O2 Cell 154
5.4 Conclusions 160
5.5 References 161
Chapter 6 Thesis Summary and Future Directions 163
6.1 Summary 163
6.2 Salt Effects on ORR 163
6.3 Solvent Effects on ORR 164
6.4 Experimental Li-Air Cells 165
6.5 Future Directions for Li-Air Research 165
Biographical Information 167
14
List of Tables
Chapter 1
Table 1.1. Standard Electrode Potentials in Aqueous Solutions at 25°C in V
vs. SHE.
Table 1.2. Practical and Theoretical Energy Capacity.
Table 1.3. Physical and Chemical properties.
Chapter 2
Table 2.1. Voltammetric properties of Fco/Fc+.
Chapter 3
Table 3.1. Physical properties of Acetonitrile.
Table 3.2. Conductivity and Viscosity of the Electrolyte Solutions in
acetonitrile.
Table 3.3. Electrochemical Charge area under the peaks. Scan rate 100mV/s.
Error ± 0.002C.
Table 3.4. Voltammetric properties of 0.1MTBAPF6 & TBAClO4 in oxygen
saturated acetonitrile. Scan rate 100mV/s. Potential error ±0.002V.
Table 3.5. Voltammetric properties of O2/O2- redox couple in 0.1MTBAPF6
& TBAClO4/MeCN.
Table 3.6. Voltammetric properties of 0.1M Li, Na &KPF6 in oxygen
saturated acetonitrile. Scan rate 25mV/s.
15
Chapter 4
Table 4.1. Conductivity of the Electrolyte Solutions.
Table 4.2. Solvent Properties.
Table 4.3. Voltammetric properties of oxygen saturated electrolytes. Scan rate
100mV/s.
Table 4.4. Oxygen Diffusion coefficient in electrolytes.
Table 4.5. O2/O2- kinetic parameters of 0.1M Li & TBAPF6.
Chapter 5
Table 5.1. Tetraethylene glycol dimethyl ether Properties.
16
List of Figures
Chapter 1
Figure 1.1. Electrochemical operation of a battery during (a) charging &
(b) discharging.
Figure 1. 2. Lithium-ion cell schematic.
Figure 1.3. Lithium-ion cell material costs.
Figure 1.4. Ragone plot showing energy density vs. power density for various
energy devices.
Figure 1.5. Lithium air cell schematic.
Figure 1.6. Electrolyte salts.
Chapter 2
Figure 2.1. Purpose built air tights electrochemical cell.
Figure 2.2. Cyclic voltammograms for the oxidation of 3.3mM Ferrocene in
1M LiPF6/1:1 EC: EMC on a glassy carbon working electrode at a scan rate of
100mVs-1.
Figure 2.3. (A) Cyclic Voltammograms for Fc/Fc+ in 1M LiPF6/1:1 EC: EMC
on a GC electrode at sweep rates between 5mVs-1 and 300 mVs-1. (B)
Randles-Sevcik plot of peak current vs. square root of the scan rate for the
curves in 2.3(A).
17
Figure 2.4. (A) Disk currents on a RDE obtained in 1M LiPF6/1:1EC: EMC
in the anodic sweep at room temperature by various rotation rates. (B) Levich
plot of limiting current vs. square root of rotation for the data in fig 2.3(A) at
scan rate = 10 mVs-1.
Figure 2.5. Tafel plots for ferrocene oxidation at room temperature on a
glassy carbon electrode at 2500 rpm for anodic sweep from 3.145V to 3.35V
at 10mVs-1 (OCP 3.145V vs Li/Li+).
Chapter 3
Figure 3.1. A) iR corrected voltammograms for the reduction of oxygen in
0.1M TBAPF6 (Black), 0.1M TBAClO4 (Blue) and the argon background
(dotted) in MeCN. B) CVs in the –2 to +0.5 V range. All scans used a glassy
carbon working electrode. Scan rate of 100mV/s.
Figure 3.2. (A) Cyclic Voltammograms for the reduction of oxygen saturated
0.1M TBAPF6 /MeCN on GC electrode at sweep rates 0.1V/s (solid), 0.1V/s
(long dash) and 0.025V/s (short dash), (B) Randles-Sevcik plot of peak
current vs. square root of the scan rate for the curves in 0.1 M TBAPF6 & 0.1
M TBAClO4/MeCN.
Figure 3.3. (A) Disk currents obtained in 0.1 M TBAPF6 MeCN during ORR
in the anodic sweep at room temperature by various rotation rates at 100mV/s.
(B) Levich plot of limiting current vs. square root of rotation in 0.1 M
TBAPF6 & 0.1 M TBAClO4 in MeCN vs. Ag/AgCl at scan rate =100mVs-1.
Figure 3.4. Tafel plots for ORR at room temperature on a glassy carbon
electrode at 2500 rpm for cathodic sweep 0.1V to -0. /s. (OCP: TBAPF6 -
0.25V & TBAClO4 -0.34 vs Ag/AgCl).
18
Figure 3.5. (A) Cyclic voltammograms of oxygen reduction in 0.1M LiPF6
(dashed line), 0.1M NaPF6 (Solid) in MeCN. Scan rate of 100mV/s (-3V to
3V vs. Ag/AgCl). (B) Oxygen reduction voltammograms in 0.1M LiPF6
/MeCN on GC electrode at various sweep rates.
Figure 3.6. Semi-log plot of potential versus log di i
i
− for the reduction of
oxygen in 0.1M LiPF6 (red), 0.1M NaPF6 (Black) and 0.1MKPF6 (Blue) in
MeCN obtained at a scan rate of 25mV/s versus Ag/AgCl .
Figure 3.7. Experimental and theoretical (n=1) √v vs. Ip plots for in 0.1M
LiPF6, NaPF6, and KPF6 & 1M KPF6 in MeCN.
Figure 3.8. Cyclic Voltammograms for the reduction of oxygen saturated 1M
XPF6 (X= Li +, Na+, K+) in MeCN on GC electrode at 500mV/s.
Figure 3.9. Steady voltammograms for the reduction of oxygen in 0.1M LiPF6
& NaPF6 in MeCN at various rotation rates at 100mV/s.
Chapter 4
Figure 4.1. Solvent Structures.
Figure 4.2. A) Cyclic voltammograms for the reduction of oxygen in 0.1M
TBAPF6 (Red, iR corrected) and the argon background (dotted) in DMSO. B)
Cyclic voltammograms (iR un-corrected) for the reduction of oxygen in 0.1M
TBAPF6/MeCN (Black), DME (Blue). Scan rate 100mV/s.
Figure 4.3. Randles-Sevcik plot of peak current vs. square root of the scan
rate in 0.1 M TBAPF6 /DMSO.
19
Figure 4.4. Levich plot of limiting current vs. square root of rotation in 0.1 M
TBAPF6/DMSO scan rate=100mVs-1 (Inset Tafel plot).
Figure 4.5. Current-voltage curves measured at 100 mV/s on a GC rotating
disk electrode (400-3600rpm) for oxygen reduction in (A) 0.1M
TBAPF6/DMSO (B) 0.1M TBAPF6/MeCN. Insets: Koutecky- Levich plot at
different potentials in kinetic-diffusion region of the polarization curve.
Figure 4.6. Cyclic voltammograms (iR corrected) for the reduction of oxygen
in 0.1M LiPF6/DMSO at various potential windows. All scans used a glassy
carbon working electrode. Scan rate of 100mV/s.
Figure 4.7. (A) Peak current vs. square root of the scan rate in 0.1 M
LiPF6/DMSO. (B) Cathodic Tafel plot obtained in 0.1 M LiPF6/DMSO
during ORR. Scan rate = 10mV/s.
Figure 4.8. Cyclic voltammograms (IR corrected) for the reduction of oxygen
in 0.1M LiPF6/MeCN at various potential windows. All scans used a glassy
carbon working electrode. Scan rate of 100mV/s.
Figure 4.9. Cyclic voltammograms (iR corrected )for the reduction of oxygen
in (A) 0.1M LiPF6/DME & (B) 0.1M LiPF6/TEGDME at various potential l
windows. All scans used a glassy carbon working electrode. Scan rate of
100mV/s.
Figure 4.10. Peak current vs. square root of the scan rate plots for the
reduction of oxygen in (A) 0.1 M TBAPF6 & 0.1 M LiPF6/MeCN. n = number
of e- (B) 0.1M TBA+ & LiPF6 /DME and 0.1M LiPF6 /TEGDME on GC
electrode.
20
Figure 4.11. Real impedance versus inverse square root of frequency in 0.1 M
LiPF6 DMSO (grey), DME (blue), TEGDME (red) and MeCN (black).
Chapter 5
Figure 5.1. Li-air cell.
Figure 5.2. Cyclic voltammograms for the reduction of oxygen in 0.1M
LiPF6/TEGDME (Blue) and the argon background (Black). Scan rate
100mV/s.
Figure 5.3. Li/air cell discharge curves at 0.25 (blue) & 0.16 (black) mA/cm2
in 1M LiPF6/TEGDME. Capacities are expressed per gram of carbon in the
electrode.
Figure 5.4. XRD pattern of fully discharged air cathode in 1M
LiPF6/TEGDME.
Figure 5.5. Full Discharge of Li/air cell discharge at 0.16mA/cm2 in 1M
LiPF6/TEGDME. Following discharge the cell was charged to 4.5V.
Figure 5.6. A) The cycling data for a 1M LiPF6/TEGDME electrolyte oxygen
cell at room temperature. The cell was discharged and charged for 2 hours at
0.13 mA/cm2. Capacities are expressed per gram of Black Pearls 2000 carbon
+ PVDF in the electrode. B) Discharge/Charge capacities as a function of
cycle number for the same cell.
Figure 5.7. A) The cycling data for a 1M LiPF6/TEGDME electrolyte oxygen
cell at room temperature. The cell was discharged and charged for 2 hours at
0.13 mA/cm2. Capacities are expressed per gram of Black Pearls 2000 carbon
21
+ PVDF in the electrode. B) Discharge/Charge capacities as a function of
cycle number for the same cell.
Figure 5.8. Discharge curves of the lithium air cell at various current
densities in 1M LiPF6/TEGDME oxygen cell at room temperature. Capacities
are expressed per gram of carbon in the electrode. (Red) 0.25mA/cm2, (Blue)
0.13mA/cm2, (Black) 0.07mA/cm2.
Figure 5.9. Nyquist impedance plots of the Li-air battery for both (9a) 2h (9b)
14h Discharge states (9c) 2h charge states at various cycles. (9)The data is
fitted by using a RC equivalent-circuit model.
Figure 5.10. Rechargeable Lithium anode.
Figure 5.11. SEM micrographs of the air cathode (11a) fresh (11b)
discharged. Scale bar is 1 µm. Energy-dispersive X-ray spectroscopy
(EDAX) (11c) fresh (11d) discharged at 0.13 mA/cm2 in oxygen.
Figure 5.12. (a) Full discharge of Li/air cell in 1M LiPF6/TEGDME (-
0.13mA/cm2). (b) Nyquist plot
22
List of Illustrations
Chapter 4
Structure I Ion pair between TBA+ and O2
- . Nitrogen is blue, carbon is gray and O is red. (Alkyl hydrogens are omitted in the structure) Structure II Ion pair between solvated Li+ and O2
- .(.The methyl hydrogens are omitted in the structure)
23
List of Abbreviations and Symbols
α
∆Ep
∆G
∆Go
ν
η
θ
ω
Å
a
A
A
b
CA
CE
CO
CV
DO
E
E0
ESR
F
Electron transfer coefficient
Peak separation
Gibbs free energy
Standard Gibbs energy
Kinematic viscosity or scan rate
Overpotential
Angle
Rotation rate
1 × 10-10 meters
Tafel intercept
Amperes
Surface area
Tafel slope
Chronoamperometry
Counter electrode
Concentration of reactant
Cyclic voltammetry
Diffusion coefficient
Electrode potential
Standard potential
Electron spin resonance
Faradays Constant
24
i
i0
ik
i lim
ipc
k0
ne
N
O
OCP
ORR
R
R
RDE
T
t
UHP
UHV
V
WE
Current density
Exchange current density
Kinetic current density
Diffusion limiting current
Cathodic peak current density
Standard rate constant
Number of electrons
Coordination number
Oxidized reactant
Open circuit potential
Oxygen reduction reaction
Reduced reactant
Gas constant
Rotating disk electrode
Temperature
Time
Ultra-high purity
Ultra-high vacuum
Volts
Working electrode
25
Chapter 1
Introduction
1.1 Energy Challenge
In mid 2009, just after taking office as Secretary of Energy, Steven Chu addressed a
clean energy forum and urged the scientific community to take on what he called "the
energy challenge." Put simply the challenge is for humans to develop alternative
energy resources that run in harmony with nature, a difficult yet not impossible task
for our highly skilled scientific community.
The search for renewable energy sources is driven by humanities appetite for
energy. By 1800s the industrial revolution was in full flight, machines were able to
replace humans resulting in higher output at lower cost. Interestingly, prior to the
industrial revolution, renewable energy resources such as wind, water and wood
supplied human energy needs. The industrial revolution led to an increase in food
production, clothing and housing resulting in unprecedented population growth 1.
This new society could now educate more individuals leading to developments in
technology and medicine in turn making life longer and more comfortable for the
masses. According to the U.N Department of Economic and Social Affairs census,
the world population has grown by 600% since the onset of the industrial revolution
and a further 200% increase is expected by 20252. Human population explosion is
inextricably linked with energy. Future population growth will drain the earth’s
resources and energy demand will increase. Renewable energy is generally an
26
inexhaustible source and typically undeveloped due to the overall reliance on fossil
fuels over the last century. Viable renewable energy sources are listed below.
o Bioenergy o Hydropower
o Electric Power o Nuclear
o Fusion o Renewables
o Geothermal o Solar
o Hydrogen o Wind
The sun delivers 800 terawatts (tW) of energy continually, of this only 18 tW
are required by a 9 billion-person planet3. A small percentage of this energy was
captured by earth and stored as non-renewable in the form of fossil fuels. We do not
have billions of years to wait for fossil fuels to be replenished. Solar energy will
likely be the game changer or a combination of these technologies. The energy
challenge at its core consists of two entwined issues. Firstly, that of energy
generation, which encompasses fossil fuels, nuclear, hydropower, wind, electric
power and so on. Secondly, once we have generated energy how do we store it?
Energy can be stored as heat in thermal storage, or as chemical energy in batteries and
capacitors. In the mid nineties Dr K.M Abraham and his colleague developed the
non-aqueous lithium air battery (Li-air)4 which addresses both these issues
simultaneously. The lithium air battery can act as both an energy source and energy
storage device. I found this energy system to be most intriguing and decided to
devote my graduate career to it.
27
1.2 Batteries
A battery allows a controlled oxidation-reduction (redox) reaction to occur to
generate electricity. Chemically energy trapped in active materials on polar
electrodes in the battery is converted to electrical energy. Electrons travel from the
negative to the positive electrode through an external circuit to power the load and
complete the discharge reaction in combination with the ions that flow between the
electrodes inside. A battery cell consists of three key components5
1) Anode (negative electrode) gives up electrons (or undergoes
oxidation reaction) during discharge.
2) Cathode (positive electrode) accepts electrons (or undergoes
reduction reaction) during discharge.
3) Electrolyte facilitates the flow of ions between the electrodes
and ultimately decides the kinetics of the reaction. The
electrolyte together with an ion-conducting separator keeps the
electrodes isolated electronically from one another to prevent
short circuit.
In a rechargeable battery the opposite processes occur during recharge (Figure 1.1).
28
Figure 1.1 Electrochemical operation of a battery during charging (a) & discharging (b)
The theoretical voltage of a reaction is determined by the difference between the
Gibbs free energy of reactants and products.
∆Goreaction = Σ ∆Go
f (products) – Σ ∆Gof (reactants) (Equation 1.1)
∆Go = nFE (Equation 1.2)
Where E is the cell voltage, n number of electrons consumed in the reaction, and F is
Faraday's constant, the charge on one mole of electrons (96500C). The capacity of a
cell is the total quantity of charge involved in a reaction defined as coulombs or
ampere-hours. Usually capacities are normalized by the mass of the material and
reported as gravimetric specific capacities, milliampere-hours per gram (mAh/g).
29
The specific energy, SE, of batteries is given in Wh/kg and is calculated by
multiplying the specific capacity by the voltage of the system.
SE (Specific Energy) = Voltage (E) x ampere-hour (Ah) (Equation 1.3)
Battery power is a function of current measured in C rates. The 1C rate is
defined as the amount of current needed to fully discharge the battery in one hour.
The lightest anode and cathode materials with the highest cell voltages lead to
the greatest energy. Table 1.1 shows standard redox potentials for various redox
couples versus a standard hydrogen electrode (SHE)6. Lithium is the strongest
reducing agent displaying the highest negative potential. Fluorine is the strongest
oxidizing agent with the largest positive potential. A Lithium-Fluorine redox couple
yields the highest theoretical voltage; unfortunately these elements react quite
violently.
Today’s battery research is focused on those based on lithium metal as it is the
most electropositive element known (-3.04V vs. SHE) and also the lightest metal (
6.94 gram per mole ). Consequently, lithium based battery systems have extremely
high energy densities
30
Table 1.1 Standard Electrode Potentials in Aqueous Solutions at 25°C in V vs. SHE.
Lithium’s unique properties high voltage, high capacity (3.86 Wh/kg) and the
ability to operate over a wide temperature range make it an ideal material for both
primary and secondary cells. Lithium anodes are quite safe for primary cells;
however safety is an issue in their rechargeable or secondary analogues. During the
charge process in a secondary lithium battery, metallic lithium is electroplated onto
the anode surface forming a porous deposit with a larger surface area than the original
metallic electrode. As the surface area increases with repeated charging and
31
discharging of the battery, metallic lithium is less thermally stable. The formation of
the high surface area lithium dendrites on the anode surface can also lead to the
shorting of cells when it grows through the separator and touches the cathode.
Removing metallic lithium and replacing it with ionic lithium (Li+) solves
many of these problems. The Lithium ion (Li-ion) concept involves replacing
metallic lithium anodes with intercalation compounds such as graphite. Carbon-
based anodes materials stabilize the electrode/electrolyte interface and can operate at
voltages outside lithium metal. Typical cathode intercalation compounds are
transition metal oxides (LiCoO2, LiMn2O4 and LiFePO4), which incorporate lithium
in their lattices and undergo oxidation to higher valences when Li is removed during
charge and vice versa when Li is inserted during discharge. Figure 1.2 shows a
schematic of a rechargeable Li-ion battery7.
Figure 1. 2 Lithium-ion schematic8.
32
Throughout the charging process lithium migrates from the cathode (for example a
lithium metal dioxide such as LiCoO2) through the electrolyte and is intercalated into
the graphite anode (LixC6). During discharge lithium is extracted from the anode and
intercalated into the cathode.
Cathode intercalation electrode materials contain transition metals (M) such as
(Co, Mn, Ni, Fe) and the prominent examples include, LiCoO2, LiNiO2,
LiNi 0.8Co0.2O2, LiMn2O4 and LiFePO4. Lithium ion batteries are the state-of-the-art
although they are not without their problems. At full charge graphite can only
incorporate one lithium per hexagon (LiC6). As a result anode capacity drops from
3861Ah/g in the case of elemental Li to 372mAh/g for graphite almost a 90%
decrease9. Internal resistance increases throughout the cycle life of the cell due to
reactions between the electrolyte and electrodes which inhibit lithium ion transport
and reduce cell capacity10,11. Battery lifetime is drastically reduced at high
temperatures as a result of increased chemical reactions at the electrode-electrolyte
interfaces12,13. Long-term storage is an issue as chemicals and materials are prone to
33
aging. In 2006 a series of fires associated with Dell laptops prompted an
unprecedented recall14 of Dell Li-ion batteries. Such incidents have highlighted the
need for safer Li-ion batteries.
Battery price is a concern especially for large batteries for electric vehicle
propulsion. One kg of oil gives about 20Wh/kg practical energy at a cost of $0.53/kg,
which is miniscule when compared to the costs of Li-ion batteries. Figure 1.3 shows
the results of a study conducted by Argonne National Lab, which found the cost of a
Li-ion battery to be around $158 per 100Ah cell having an energy density of
100Wh/kg.
Figure 1.3 Lithium-ion cell material costs15.
Expensive cathodes are about 50% of the cost for these cells, twice as much as
other components. Research efforts are focused on the development of low cost long
life materials, particularly electrode materials, and the key to lowering the price of
rechargeable Li batteries. In this respect Li-air battery is highly promising as the
electroactive element O2 is free and environmentally friendly.
34
1.3 Fundamental of the Lithium-Air Battery
The Lithium-air battery is one of the most energy dense electrochemical power
sources. Table 1.2 compares theoretical energy capacities of metal air batteries to
well established systems.
Table 1.2 Practical and Theoretical Energy Capacities.
35
Figure 1.4 Ragone plot showing energy density vs. power density for various energy devices.
The Ragone plot (fig 1.4) shows Li-air has both high energy and power
densities. Oxygen the cathode active material is not stored in the battery but is
accessed from the environment, and in turn significantly reduces the battery’s total
mass. The Li/O2 redox couple has the highest theoretical energy density of any
known viable redox couple and has the potential to significantly increase the energy
density of practical batteries. Fully developed and optimally packaged Li-air batteries
could exceed specific energies of 2000 Wh/kg, table 1.2 shows only gasoline has a
higher theoretical energy density. However at only 20% efficiency its practical
energy density pales in comparison to the Li-air battery. Specific energies for the
metal-air cells were calculated using Gibbs energies of formation according to
∆Goreaction = Σ ∆Go
f (products) – Σ ∆Gof (reactants)
36
In practice, oxygen is not stored in the battery, which increases the theoretical
specific capacity of the cell to 11,140 Wh/kg although the battery weight would
increase as battery is discharged. Figure 1.5 shows a schematic of a Li-air cell.
Lithium metal is oxidized at the anode to form Li+ ions, which migrate towards the
cathode. The electrons from the oxidation reaction at the anode are passed through an
external circuit to perform work in the load, and are returned to the cell at the cathode
to complete the electrochemical reaction in combination with the Li ions that migrate
to it from the anode. Oxygen is reduced at the cathode in the presence of the supplied
electrons and Li+ ions to one or more of following products, LiO2, Li2O2 and Li2O.
Figure 1.5 Lithium air cell schematic.
37
The possible reactions of the Li-air cell at the cathode and anode are:
Cathode
(1) O2(g) + Li++ e → LiO2(s) ∆Go = -70 kcal (Eo= 3.0V)
(2) O2(g) + 2Li++ 2e- → Li2O2(s) ∆Go = -145 kcal (Eo = 3.1 V)
(3) 2O2(g) + 4Li++ 4e -→ 2Li2O(s) ∆Go = -268 kcal (Eo = 2.91 V)
Anode
Li(s) → Li++ e- (Eo = 0.00 V)
Recently, we have shown16 that the first product of the reduction of oxygen in
non-aqueous electrolytes is superoxide, O2-, involving a one-electron process. We
also found that the half-life of the superoxide depends on the cation present in the
electrolyte solution. The most prominent product in Li-air cell as discussed in
Chapters 3-5 is Li2O2.
1.4 Lithium-Air Today
The non-aqueous Lithium air battery is a relatively new technology. It all began in
19964 when a battery technician accidently introduced a little oxygen to a Li/graphite
half-cell of a Li-ion battery. Slightly bemused by the large increase in cell voltage, he
showed his results to his boss (Dr K.M Abraham). He quickly recognized the
importance of this accidental observation and put it together by devising a series of
experiments which led to his seminal paper and the introduction of the non-aqueous
organic Li-air battery. In the 15 years since, Li-air has emerged as major candidate
38
for alternative energy in the future. This first paper was quite bold in that it addressed
the major drawbacks of the Li- air battery listed below.
a) Oxygen Solubility
b) Lithium Oxide dissolution
c) Stability of Lithium anode
d) Catalyst Development for rechargeability
Rechargeability is the most significant obstacle that has to be overcome before
full capability of the battery is realized as a renewable energy storage system17.
Oxidation of the reduction products is thermodynamically unfavorable and suffers
from poor kinetics. The first Li-air cell was composed of a Li anode, a
polyacrylonitrile-based gel polymer electrolyte and a porous carbon cathode. In the
absence of a catalyst the oxidation reaction occurs near 4V, and a large hysteresis
between charge and discharge voltages was observed. The hysteresis was reduced by
employing a cobalt phthalocyanine (CoPc) - based oxidation catalyst which also
improved charge/discharge efficiency. Recent investigations have employed
manganese oxide (MnO2) catalysts18,19 although the charge voltages in these cells are
similar to the uncatalyzed cells. Through Raman spectroscopy Lithium peroxide
(Li 2O2) was identified as the chief discharge product5. The formation of Lithium
peroxide is consistent with the open circuit voltage (OCV) of about 2.9V measured
for the cell and the theoretical voltages calculated for the reactions in equations 1-3.
Li-air cells shares similar drawbacks relevant to both Li-ion and fuel cell
technologies, therefore teething problems were surmounted by applying known
39
solutions. An avenue of investigation is applying existing electrolytes from
conventional Li-ion batteries to Li-air. Initially Jeffery Read investigated possible
electrolytes, by drawing on his experience with Li-ion electrolytes20-22. The results of
his studies found electrolyte formulation has a large influence on cell performance.
Kuboki et al23 studied the performance of hydrophobic ionic liquids in an ambient
environment as electrolytes. Ionic liquids demonstrated high lithium stability and
high discharge capacities. A number of groups are interested incorporating solid
electrolytes in Li-air batteries. Protected lithium electrodes (PLE) such as Lisicon24
have been applied successfully in both aqueous and non-aqueous Lithium batteries.
Such coating protect against moisture permeation into the cell especially to the anode.
Low carbon loading on nickel foam25 has demonstrated the highest discharge capacity
thus far (5,000mAh/g).
Li-air discharge products are fairly insoluble in contemporary Li-ion organic
electrolytes. Rechargeability maybe enhanced by suitable organic solvents.
Electrode structure is crucial as it sets the stage for the oxygen reduction reaction
(ORR). Appropriate electrode morphology, surface structure, pore volume and
surface area can enhance the rechargeability of the Li-air cell. Our recent studies
have revealed that the Li/O2 cell can be recharged with high efficiency without a
catalyst by using appropriate porous carbon electrodes. Interestingly charge voltages
of these uncatalyzed cells are similar to those of the MnO2 catalyzed cells with both
of these cell exhibiting higher charge voltages than the cobalt-catalyzed cells. Studies
discussed in Chapter 5 have uncovered factors limiting the rechargeability of the Li-
air battery
40
1.5 Non- Aqueous Electrolytes
The majority of electrochemical reactions are carried out in solution. A liquid
medium allows control of reaction conditions, i.e., temperature, pressure, rate of mass
transfer and reactant concentration. Water is the most popular solvent, its high
polarity (78 ε (dielectric constant)) make it ideal for dissolving a wide variety of salts.
In certain circumstances water maybe an undesirable medium for the following
reasons:
1) Water is a source of protons, which are highly reactive with electrode
materials and alkali earth metals, which undergo fast hydrolysis.
2) The electrochemical window of water is too narrow (1.299V). Hydrogen and
oxygen evolution occurs at cathode and anode, respectively. Numerous
electrochemical reactions of high energy density batteries occur beyond these
voltage limits.
3) Many chemical compounds are insoluble in water
4) Aqueous electrolytes are limited by temperature, i.e. the boiling point of water
of 100oC, which is too low for many practical uses in conversion and energy
storage.
Solvents other than water are generally called non-aqueous solvents. Appropriate
non-aqueous solvents can dissolve substances that are insoluble in water, stabilize
substances that are unstable in water, and facilitate electrochemical reactions that are
otherwise impossible. The electrochemical window of non-aqueous solvents is much
larger than in water. As a result the field of non-aqueous electrochemistry has
41
attracted increasing interest for energy storage. Electrolyte consists of solvents such
as molecular liquids or ionic liquids that dissolve solutes, which can be solid, liquid
or gaseous. Electrolytes can be liquid solutions based on organic and inorganic
solvents or molten salts. Solid electrolytes such as ionically conducting polymers and
conducting solids such as doped oxides and glasses have a whole host of applications.
Non-aqueous liquid electrolytes may be divided into protic or polar aprotic
solvents. Protic solvents donate protons (H+), solvents containing amine or hydroxyl
groups are protic. These solvents generally have low dielectric constants and low
polarity and are reactive toward electrode materials, particularly Li. Aprotic solvents
do not contain acidic hydrogen’s. In this work our attention will be devoted to polar
aprotic solvents as they are the most important and useful with respect to the Li-air
battery. Common organic carbonates, esters and ethers used as solvents in lithium
chemistry are shown in table 1.3.
42
Table 1.3 Physical and chemical properties26
43
These solvents are useful for lithium battery applications for the following reasons;
1) Wide electrochemical window
2) Low volatility
3) Low reactivity towards electroactive species and electrode materials
4) High polarity (Ability to dissolve salts especially Li salts)
Formulating the ideal electrolyte is about finding the right blend of both physical
and chemical properties. The boiling point of the solvent is crucial to an
electrochemical experiment conducted at both low and high temperatures. Solvent
viscosity strongly influences mass transport of electroactive species hence
determining kinetics of the reaction. The dielectric constant (εr) is measure of
polarity of the solvent to enable salt dissociations in to ions. Polar solvents separate
charged particles, by weakening their respective electrostatic forces. The ideal
electrolyte for the Lithium air battery requires solvents of low viscosity, high
dielectric constant, high oxygen solubility and low water solubility.
Lithium salt criteria:
1) Lithium salts should be able to completely dissolve and dissociate in the non-
aqueous solvent.
2) The anion should be inert and stable to the cathode potential.
3) Anion and cations should be inert to cell components such as electrode
material and separators.
4) Remain thermally stable
44
Although the availability of non-aqueous solvents is large, the choice of lithium salts
is quite limited. Ions of low charge density usually lead to good charge solubility and
separation if paired with bulky anions and or cations. Lithium is often paired with
bulky anions such as I-, Br- ClO4-, PF6
-, BF4-, RCO2
-. Most Lithium salts are rather
difficult to dissolve due to the small ionic radius of Li+. Eligible candidates are
usually based on large anions such as PF6- where essentially F- is stabilized by PF5 by
distributing the charge throughout the ion. The corresponding lithium salts (LiPF6)
usually are better dissociated and solvated in low dielectric solvents. For
fundamental electrochemistry using a large like Tetra butyl ammonium (TBA+)
eliminates complications associated with Li+ such as insoluble salt precipitates. The
solvating power of solvent is complex especially in electrochemistry. Although
dielectric constant is the primary measure of polarity other factors such as acidity,
basicity and structure are crucial.
To understand the solvation of metal cations one must understand acid base
chemistry. According to this theory metal cations act as Lewis acids and solvent
molecules act as Lewis bases. Lewis acids act as electron pair acceptors and Lewis
bases electron donors. Gutmann27 developed the acceptor number (AN) and the
donor number (DN) model as a measure of a solvents Lewis acidity and Lewis
basicity respectively. The higher the DN or AN of a solvent the stronger its basic or
acidic character.
45
1.6 Non-Aqueous Oxygen Reduction Reaction (ORR)
Oxygen is the essence of all life; and we are only beginning to understand its role in
nature. In biological terms oxygen is reduced by cellular respiration, enzyme reaction
and photochemistry to produce oxygen radicals known as superoxides (O2-). In its
free radical state this ion can be detrimental to tissue and DNA28. However harmful
these free radicals maybe biologically, their chemistry is proving very attractive as
energy sources. The electrochemical reduction of oxygen to superoxide can be taken
advantage of as the superoxide ion can behave as a Lewis base, nucleophile, as a well
as both an oxidizing and reducing agents.
These traits make the reduction of oxygen ideal for energy production and
storage. Unfortunately superoxide has a relatively short lifetime in aqueous media
due to the presence of protons, which result in the direct 2e- reduction of oxygen to
hydrogen peroxide.
Nonaqueous solvents free of proton interference are an ideal environment for
stable oxygen reduction. Electrochemical studies in non-aqueous oxygen reduction
reaction (ORR) begun in the early 1960’s. The original research utilized aprotic
organic solvents such as dimethylsulfoxide (DMSO), dimethylformamide (DMF) and
acetonitrile (MeCN)29-32 and bulky salts like tetraalkylammonium perchlorate
(NR4+ClO4) for the reasons listed above. This early work demonstrated for the first
time a reversible oxygen redox couple stable in aprotic solvents based on quaternary
46
ammonia salts. More recently stability of ionic liquids towards superoxide has been
examined33-35. We seek to expand our understanding of the electrochemical behavior
of oxygen in aprotic media in presence of Li salts, as they are relevant to the Li-air
battery.
The work as discussed in this thesis shows that the electrochemistry of oxygen
in non-aqueous electrolytes in presence of alkali metal salts is markedly different
from that in the presence of the akyl ammonium salts and that in aqueous media. We
are interested in using small cations of lithium (Li), sodium (Na) and potassium (K)
in an effort to apply oxygen chemistry towards metal-air batteries. We carried out
this study in a series of electrolyte solutions of hexafluorophosphate salts (X+ PF6-),
(X = TBA, K, Na, Li) shown in figure 1.6.
Figure 1.6: Electrolyte salts. This thesis reports our fundamental studies aimed at rationally designing electrolytes
for the Li-air battery.
47
1.7 Scope of Dissertation
At the outset of this work we sought to fully elucidate the oxygen reduction reactions
(ORR) in the non-aqueous environment. These studies of ORR were performed using
standard electrochemical techniques such as cyclic voltammetry (CV), rotating disk
electrode (RDE) and electrochemical impedance spectroscopy. The studies of Li-air
cell charge and discharge were combined with standard analytical techniques such as
X-ray diffractometry to analyze discharge products.
The focus of chapter 2 is two-fold. First we wished to evaluate and validate
our experimental techniques for the ORR studies using a well-known redox couple
for which the ferrocene/ferrocenium (Fc/Fc+) redox couple is a prominent one for use
in non-aqueous electrolytes. Coincident with this was the second aspect of the studies
involving the Fc/Fc+ couple as a redox reagent for the overcharge protection of
rechargeable Li batteries. Such reagents are necessary for building high voltage
batteries from single Li-cells by connecting them in series.
In Chapter 3 we investigate the effects of conducting salts (anion and cation)
on the ORR in acetonitrile. We found that the conducting salt significantly affected
the reversibility and kinetics of oxygen reduction in non-aqueous electrolytes. The
effects of organic solvents on ORR are examined in chapter 4. The results of this
study show that the solvent and the supporting electrolyte act collectively to influence
the nature of reduction products and their rechargeability. In chapter 5 we pool all the
knowledge collected throughout these ORR studies to construct and discharge and
charge the first uncatalyzed Li-air battery.
48
The significance of this thesis is the insight gained into non-aqueous ORR which can
be directly applied to the Li-air battery. The legacy of this work is to provide a
methodology for rationally designing and selecting appropriate electrolytes for Li-air
batteries and developing a fundamental mechanism for ORR in non-aqueous
electrolytes.
1.8 Chapter 1 References
(1) Mokyr, J. The British Industrial Revolution: An Economic Prespective; 2nd ed.; Westview Press, 1999. (2) Division, P.; D.E.S.A, Ed.; U.N: Vienna, 2008. (3) Lewis, N. S.; Nocera, D. G. Proceedings of the National Academy of Sciences 2006, 103, 15729-15735. (4) Abraham, K. M.; Jiang, Z. J. Electrochem. Soc 1996, 143, 1-5. (5) Linden, D.; Reddy, T. Handbook of Batteries; Third ed.; McGraw- Hill, 2001. (6) A.J.Bard Electrochemical Methods Fundamentals and Applications; 2 ed.; John Wiley & Sons: New York, 2001; Vol. (7) Abraham, K. M. Journal of Power Sources 1985, 16, 171-178. (8) Tarascon, J. M.; Armand, M. Nature 2001, 414, 359-367. (9) Zaghib, K.; Nadeau, G.; Kinoshita, K. Journal of The Electrochemical Society 2000, 147, 2110-2115. (10) Abraham, K. M.; Foos, J. S.; Goldman, J. L. Journal of The Electrochemical Society 1984, 131, 2197-2199. (11) Broussely, M.; Herreyre, S.; Biensan, P.; Kasztejna, P.; Nechev, K.; Staniewicz, R. J. Journal of Power Sources 2001, 97-98, 13-21. (12) Abraham, K. M.; Harris, P. B.; Natwig, D. L. Journal of The Electrochemical Society 1983, 130, 2309-2314.
49
(13) Winter, M.; Brodd, R. J. Chemical Reviews 2004, 104, 4245-4270. (14) Dell In Battery Recall 2006. (15) Gaines, L.; Cuenza, R. Costs of Lithium-Ion-Batteries for Vehicles Argonne National Laboratory, 2000. (16) O 'Laoire, C.; Mukerjee, S.; Abraham, K. M.; Plichta, E. J.; Hendrickson, M. A. The Journal of Physical Chemistry C 2009, 113, 20127-20134. (17) Abraham, K. M. ECS Transactions 2008, 3, 67-71. (18) Dobley, A.; Morein, C.; Abraham, K. M. ECS Meeting Abstracts 2006, 502, 823-823. (19) Débart, A.; Paterson, Allan J.; Bao, J.; Bruce, Peter G. Angewandte Chemie International Edition 2008, 47, 4521-4524. (20) Read, J. Journal of The Electrochemical Society 2002, 149, A1190- A1195. (21) Read, J.; Mutolo, K.; Ervin, M.; Behl, W.; Wolfenstine, J.; Driedger, A.; Foster, D. Journal of The Electrochemical Society 2003, 150, A1351-A1356. (22) Read, J. Journal of The Electrochemical Society 2006, 153, A96- A100. (23) Kuboki, T.; Okuyama, T.; Ohsaki, T.; Takami, N. Journal of Power Sources 2005, 146, 766-769. (24) Wang, Y.; Zhou, H. Journal of Power Sources 2009, 195, 358-361. (25) Beattie, S. D.; Manolescu, D. M.; Blair, S. L. Journal of The Electrochemical Society 2009, 156, A44-A47. (26) Xu, K. Chemical Reviews 2004, 104, 4303-4418. (27) Gutmann, V. Coordination Chemistry Reviews 1976, 18, 225-255. (28) Chin, D. H.; Goldberg, I. H. Biochemistry 1986, 25, 1009-1015. (29) Maricle, D. L.; Hodgson, W. G. Anal. Chem. 1965, 37, 1562-1565. (30) Peover, M. E.; White, B. S. Electrochimica Acta 1966, 11, 1061-1067.
50
(31) Sawyer, D. T.; Roberts, J. L. J. Electroanal. Chem. 1966, 12, 90-101. (32) Johnson, E. L.; Pool, K. H.; Hamm, R. E. Anal. Chem. 1967, 39, 888- 891. (33) Buzzeo, M. C.; Klymenko, O. V.; Wadhawan, J. D.; Hardacre, C.; Seddon, K. R.; Compton, R. G. The Journal of Physical Chemistry A 2003, 107, 8872-8878. (34) Carter, M. T.; Hussey, C. L.; Strubinger, S. K. D.; Osteryoung, R. A. Inorganic Chemistry 1991, 30, 1149-1151. (35) Huang, X.-J.; Rogers, E. I.; Hardacre, C.; Compton, R. G. The Journal of Physical Chemistry B 2009, 113, 8953-8959.
51
Chapter 2
Electrochemical Studies of Ferrocene in a Lithium Ion
Conducting Organic Carbonate Electrolyte
2.1 Introduction
Interest in non-aqueous solvents for electrochemical research and practical
applications such as lithium batteries has increased significantly over the past four
decades.
Ferrocene (Fco) is a useful reference material for non-aqueous
electrochemistry as it demonstrates good solubility, invariant redox potentials and
excellent chemical and electrochemical reversibility in organic electrolytes1. The
reversibility of the (Fco/Fc+) redox couple was established from polarographic
studies2 soon after the discovery of this organo-iron compound in 1951 by Kealy and
Pauson3. Previous studies4 of the electrochemistry of ferrocene in various non-
aqueous solvents revealed a reversible one-electron process.
The diffusion coefficient (D) of ferrocene in different solvents was found to
be inversely dependent on the viscosity of the solvent medium. Weaver et al5
conducted a study of the thermodynamic effects of solvent dynamics on various
metallocene redox couples by both theoretical and experimental methods. Their
results indicated that solvent viscosity contributed to the high energy barrier, which
influenced the kinetics of outer-sphere reactions. Mass transfer of the electroactive
species to the electrode surface is a major factor in the rate of an electrochemical
52
oxidation or reduction reaction. If the electron transfer step is not hindered
kinetically, movement of the electroactive species through the solution becomes the
rate-limiting step in this case. As a result, electrochemical measurements are
frequently used to determine diffusion coefficients of electroactive species and
kinetics of electrode reactions. The process of diffusion is important in a wide variety
of chemical scenarios, including kinetics of rapid reactions, chromatographic and
electrophoretic separations, and battery electrode reactions.
A literature review6-10 revealed some prior studies of ferrocene
electrochemistry in propylene carbonate (PC) solutions containing lithium salts. In
the first of these studies, Abraham et al10 investigated the electrochemical properties
of Fc in a polyacrylonitrile-based gel polymer electrolyte (PAN)-EC/PC-LiC1O4) and
established that the oxidation of ferrocene is electrochemically reversible. They
found that the diffusion coefficient of Fc decreased by an order of magnitude in the
gel polymer electrolyte compared with liquid electrolytes having similar Li salt
concentrations.
To the best of our knowledge, few studies concentrating on ferrocene
oxidation kinetics in highly concentrated solutions of Li salts in organic carbonates of
the types used in Li-ion batteries have been performed. Such studies are relevant in
view of the fact that ferrocene and its derivatives have been shown to be potentially
useful redox reagents for the chemical overcharge protection of rechargeable lithium
and lithium-ion (Li-ion) batteries11-15. In this application, ferrocene added to the
electrolyte in a rechargeable Li or Li-ion battery cell is oxidized at a potential slightly
positive of the oxidation potential of the positive electrode in the cell and the
53
ferrocenium ions (Fc+) thus produced diffuse to its negative electrode and gets
reduced to regenerate ferrocene. Consequently, the electrode potential remains
locked at the oxidation potential of ferrocene and prevents the cell from overcharge.
This type of chemical shuttles for overcharge protection is highly desirable to protect
individual cells in a battery having two or more cells connected in series from
overcharge during recharge with the result of improving cell performance and
mitigating safety hazards.
In this work, we investigated ferrocene redox chemistry in 1M LiPF6/1:1 EC:
EMC which is a typical liquid electrolyte used in Li-ion batteries. Another motivation
for our study is that these ferrocene experiments can serve as models for investigating
the redox chemistry of other chemical shuttle reagents used for overcharge protection
of rechargeable Li-ion batteries16-19. We report the mass transport and kinetic
parameters of the Fco/Fc+ couple in this prototypical Li-ion battery electrolyte. The
results of this study should further our ability to design and develop redox shuttles in
non-aqueous electrolytes leading to improved performance and safety in Li-ion and
Li–air batteries20. A comparison of the results obtained from CV and RDE
experiments is useful in understanding the role of mass transport on the kinetic
parameters of redox reagents for Li-ion batteries.
2.2 Experimental
2.2.1 Chemical Reagents
All reagents were electrochemical grade unless stated otherwise stated. Battery grade
solvents, Ethylene Carbonate (EC) and Ethyl Methyl Carbonate (EMC) and Lithium
54
hexafluorophosphate (LiPF6) (battery grade, >99.9%, H2O< 20ppm) were obtained
from Ferro Corporation, Cleveland, Ohio. Ferrocene was purchased from Sigma-
Aldrich, Allentown, PA.
2.2.2 Instrumentation
The electrochemical experiments were performed with a VoltaLab (Radiometer
Analytical Inc, model-VoltaLab 10) potentiostat in an air-tight electrochemical cell.
Figure 2.1 show’s the electrochemical cell which was designed and built in-house. It
consisted of a traditional 3-electrode system utilizing Li/Li+ as the reference electrode
and platinum wire as the counter electrode.
Figure 2.1: Purpose built air tight electrochemical cell
A glassy carbon working electrode (3mm diameter) was employed for the cyclic
voltammetry experiments. The electrodes were polished with 0.5 and 0.05 mm
55
alumina paste prior to the experiments. For RDE experiments, the glassy carbon
electrode was rotated with an Autolab RDE rotor. Scan rate analyses were performed
using 3.3mM solutions of ferrocene in a 1MLiPF6/1:1(by volume) EC: EMC
electrolyte, and scan rates were varied between 5mV/s and 300mV/s. All of the
cyclic voltammetry experiments were initially performed in an argon (Ar)-
atmosphere glove box where H2O and O2 concentrations were kept below 5ppm and
temperature was held at 22 ± 2°C. RDE experiments were conducted outside the
glove box in a glove bag purged with argon.
The conductivities of the solutions were measured using a 4-electrode
conductivity cell with a Thermo Scientific Orion Model 550A Multiparameter Meter.
Electrolyte viscosities were measured with a size 1 Ubbelohde Viscometer.
Measurements and calibration were performed according to ASTM protocol as
described by the manufacturer. Efflux time between the upper and lower fiducial
marks on the apparatus were monitored with a stopwatch. Average times for four
runs were recorded, with all results averaged together. The viscosities were measured
at room temperature, 22± 2°C.
2.3 Results and Discussion
2.3.1 Cyclic voltammetry
The redox chemistry of ferrocene was studied using cyclic and rotating disc
voltammetric techniques and the results are compared. The electrolyte solution used
for these studies was characterized by determining its conductivity and dynamic
viscosity with values of 8.8 mS/cm and 4.66cP, respectively at room temperature.
Figure 2.2 shows the cyclic voltammograms (CVs) of the ferrocene redox couple on a
56
glassy carbon (GC) electrode in 1MLiPF6/1:1EC:EMC solution (hereafter referred to
also as the carbonate electrolyte) at a sweep rate of 100 mVs-1.
2.8 3.0 3.2 3.4 3.6
-2e-5
-1e-5
0
1e-5
2e-5
3e-5
ExperimentalIR Corrected
Cur
rent
(A c
m-2
)
Potential (V vs.Li/Li +)
Figure 2.2: Cyclic voltammograms for the oxidation of 3.3mM Ferrocene in 1M LiPF6/1:1 EC: EMC on a glassy carbon working electrode at a scan rate of 100mVs-1.
Our study was restricted to glassy carbon electrodes, since surface adsorption
effects on platinum electrodes are well documented 21. Also, glassy carbon electrodes
are practically more relevant to Li-ion batteries as one or another form of carbon is
present in the electrodes of the battery. At low concentrations, ferrocene oxidation in
many organic electrolyte solutions does not precipitate surface films on glassy carbon
electrodes and, thus, would serve as a good standard for non-aqueous
electrochemistry. The CV data were corrected for Ohmic (iR) losses using the well
established semi-integral technique22, (see fig 2.2). Equilibrium is established
57
quickly between the active species as the voltammetric responses of ferrocene appear
between 3.22 and 3.28V vs. Li/Li+. The peak potential separation ∆Ep between the
anodic and cathodic peak potentials ranged from 60-67mV with an average of 63 ±
0.002 mV. These values are close to the theoretical value of 59mV for a one-electron
reaction. For a reversible process the peak width is given by the following
relationship.
/ 2 - 2 .2p pR T
E En F
=
(Equation 2.1)
Where Ep/2 is the half-peak potential at the half value of the peak current, Ep, is the
peak potential, F is the Faraday constant and n is the number of electrons in the
reaction. From the data obtained at the sweep rate of 5mV/s, the number of electrons
n was calculated to be 1.05 (see Table 2.1).
58
Analysis of the CVs over the whole sweep ranges (see Table 2.1) gave n
values close to this indicating that the number of electrons transferred in the reaction
is one. We found the electrochemical charge ratio (Qc/Qa) determined from the area
under the oxidation (ipa) and reduction (ipc) peaks to be over 91 %.
Cyclic voltammetry is a useful technique for discerning kinetics, rates, and
mechanisms in addition to thermodynamic parameters. The magnitude of the current,
I, in a cyclic voltammogram is a function of temperature, concentration, Canalyte,
electrode area, A, the number of electrons transferred, n, the diffusion coefficient, D,
and the speed at which the potential is scanned, v, related by the Randles-Sevcik
equation (Equation 2.2).
5 3 2 1 2 1 2paI =(2.69×10 )n AD v C (Equation 2.2)
Figure 2.3a shows both the cathodic and anodic peak potential variations with
sweep rates. The dependence of Ipa on sweep rate is evident at both low and high
scan rates. The little variation in both the reduction and oxidation peak positions with
increase in sweep rates reflects the reversibility of the system. The Randles–Sevick
plot (Fig 2.3b) shows a linear relationship between Ip vs. v1/2 passing through the
origin. Asumming n=1 we calculated the theoretical Randles-Sevick plots, which
agree with the experimental data. This infers that ferrocene is a reversible redox
couple in this media and follows scheme 2.1.
( ) ( ) −++ +⇔ eHCFeHCFe 2553
2552 (Scheme 2.1)
59
2.8 3.0 3.2 3.4 3.6
Cur
rent
(A c
m-2
)
-4e-5
-2e-5
0
2e-5
4e-5
5mV/s25mV/s 50mV/s75mV/s 100mV/s 200mV/s300mV/s
0.0 0.1 0.2 0.3 0.4 0.5 0.6-8e-5
-6e-5
-4e-5
-2e-5
0
2e-5
4e-5
6e-5
8e-5
ExperimentalTheoretical(n=1)
V1/2(v/s)1/2
B
A
Potential (V vs.Li/Li +)
Figure 2.3: (A) Cyclic Voltammograms for Fc/Fc+ in 1M LiPF6/1:1 EC: EMC on a GC electrode at sweep rates between 5mVs-1 and 300 mVs-1. (B) Randles-Sevcik plot of peak current vs. square root of the scan rate for the curves in 2.3(A).
60
The voltammetric parameters for 3.3mM Fc in the carbonate electrolyte are
summarized in Table 2.1. All potential values are reported versus the Li/Li+
reference electrode. From the dependence of current on scan rate, the diffusion
coefficients of ferrocene and ferrocenium ion in solution were calculated as 2.03 x
10-6 ± 0.02 cm2 sec -1 and 1.38 x 10-6 ± 0.01 cm2 sec -1 , respectively. The lower
diffusion coefficient of ferrocenium ion is understood as arising from interactions
between the ferrocenium cation and the PF6- anion as well as solvent molecules. The
diffusion coefficient of ferrocene we found is an order of magnitude lower than that
reported for PC/0.1M LiClO46. The lower value obtained in the present work is
attributed to higher viscosity of the concentrated solution. This points out the
importance of the present work.
The diffusion coefficient of a species in solution is inversely proportional to
the viscosity of the solution according to Walden’s Rule. The theoretical diffusion
coefficient of this system can be determined by the relationship between diffusion
coefficient and solution viscosity given by the Stokes Einstein equation (Equation
2.3).
kTD=
6πηa (Equation 2.3)
In this equation k is the Boltzmann constant, T is temperature, η is dynamic viscosity
and a is the effective hydrodynamic radius of ferrocene. The hydrodynamic radius is
influenced by a number factors such as solubility of analyte, solvent molecule size
and polarity of the solvents. In this case we decided to use the crystallographic radius
of ferrocene (0.32nm)23. The dynamic viscosity (4.65cP) was calculated from the
solution’s kinematic viscosity (0.0372 cm2s-1) and density (1.25gcm-3) are measured.
61
Potential(V)
2.8 3.0 3.2 3.4 3.6 3.8
0
2e-5
4e-5
6e-5
2 4 6 8 10 12 14 16 180.0
2.0e-5
4.0e-5
6.0e-5
8.0e-5
1.0e-4
1.2e-4
1.4e-4
ExperimentalTheoretical n=1 Theoretical n=2
Cur
rent
(Am
ps)
ωB
A
2025rpm
2500rpm
1600rpm
1225rpm
900rpm
625rpm
400rpm
Figure 2.4: (A) Disk currents on a RDE obtained in 1M LiPF6/1:1EC: EMC in the anodic sweep at room temperature by various rotation rates. (B) Levich plot of limiting current vs. square root of rotation for the data in fig 2.3(A) at scan rate = 10 mVs-1.
62
The theoretical diffusion coefficient is D = 1.50 x 10-6 cm2s-1, very close to the
measured value.
2.3.2 Rotating Disk Electrode
Rotating disk electrode is a hydrodynamic electrode technique which utilizes
convection as the mode of mass transport as opposed to CV which is governed by
diffusion. Convection is more efficient and is not diffusion limited with the result
that the analytical data is more reproducible and precise. Thus a comparison of the
kinetic parameters obtained from CV and RDE experiments is informative to
elucidate the role of mass transport on electrode reaction kinetics. Figure 2.4a shows
RDE voltammograms for ferrocene at a series of rotation rates. It is evident from the
data that the current generated by the RDE method is much larger than that generated
under diffusion control (Figure 2.1a). The much larger current obtained using RDE
reflects the efficiency of this method. Also notice that there is significant increase in
anodic current (i.e.Fc0 to Fc+) while the amount of cathodic current (i.e. Fc+
to Fc0) is
negligible, essentially making the cyclic voltammogram anodic. This is due to the
vast difference in concentration between the Fc+ and Fc0. The bulk solution contains
Fc, which provides a constant supply to the rotating electrode while the concentration
of Fc+ ions at the electrode is so minuscule that little anodic current is produced. The
Levich equation (equation 4.4) establishes relationship between current at the RDE
and concentration of the analyte.
CvωD(0.620)nFAI 1/6212/3lim
−/= (Equation 4. 4)
63
, where Ilim is the limiting current density (Acm-2), n is the number of electrons for the
reaction, F is the Faraday constant (96,500 Cmol-1), D is the diffusion coefficient of
ferrocene in the solution, v is the experimentally determined kinematic viscosity of
the solution (0.0372cm2s-1), C is the concentration of ferrocene in the solution
(3.3mM) and ω is the angular frequency (2
60
fπ). Furthermore, the Levich equation
allows us to construct a plot of Ilim versus ω1/2 to determine a value of D which was
found to be 2.35 x 10-6 ± 0.2 cm2 sec -1. This value is slightly higher that determined
from the aforementioned CV experiments. RDE provides insight into the number of
electrons transferred in the electrochemical reaction by comparing the limiting
currents to the rotation rate of the electrode. These Levich plots, shown in figure
2.3b, display well defined linear plots indicating a simple mass transfer controlled
electrode process. The slope of the Levich plot for the experimental data closely
parallels the theoretical line for a one electron reaction (n=1).
We can apply the Tafel equation (equation 2.5) which relates the rate of an
electrochemical reaction to the overpotential, according to
1
log log onF
i iRT
αη
− = +
(Equation 2.5)
In this equation η is overpotential for the anodic reaction and the other symbols have
their usual meaning. The Tafel plot is corrected for diffusion. In order to correct the
measured currents for diffusion, the kinetic current in the mixed activation-diffusion
region is calculated from equation (2.6) 24.
64
lim
lim
.
k
i ii
i i=
− (Equation 2.6)
A plot of log ik against overpotential, η, should be linear, leading to the Tafel slope b,
from which the transfer coefficient, α can be determined. As already stated (Fig 2.4)
the RDE experiments are related exclusively to the oxidation of ferrocene. Figure 2.5
shows the Tafel plot from the anodic region of the voltammogram beginning with the
open circuit potential (OCV) of 3.145. In this plot, the Tafel region starting at about
60 mV positive of the OCP is clearly delineated from the Butler Volmer region below
that. A single Tafel slope of c.a. 79mV/decade was obtained in the entire potential
range for all rotation rates. This slope is analogous that obtained by Petrocelli 25 for
the oxidation of Potassium Ferricyanide in NaOH on platinum. This implies that the
initial electron transfer is the rate-limiting step of this reaction.
65
0.00 0.05 0.10 0.15 0.20 0.25-8
-7
-6
-5
-4
-3
-2lo
g i k
(A
cm
-2)
η η η η (V vs.Li/Li +)
Figure 2.5: Tafel plots for ferrocene oxidation at room temperature on a glassy carbon electrode at 2500 rpm for anodic sweep from 3.145V to 3.35V at 10mVs-1 (OCP 3.145V vs Li/Li+).
The heterogeneous kinetics of this reaction is so rapid that over a wide range
of sweep rates, the reaction is reversible. From this data we calculated the transfer
coefficient α = 0.3, which is comparable to previous ferrocene experiments in aprotic
solvents26,27. The observed values of Tafel slope and α are indicative of strong
interactions between the ferrocenium ions and PF6- as well as the solvents. The low α
also suggests that the structure of the activated complex for the oxidation reaction is
closer to that of the oxidized specie. As we noted earlier about 8% of ferrocenium
ions are not available for reduction back to ferrocene. This together with the kinetic
information suggest a chemical step, following the one electron rate determining
66
oxidation reaction, in which the ferrocenium ion formed is stabilized by the solvent as
well as some of it being transformed into products. This is not unreasonable
considering the dipolar nature of the organic carbonate solvents. Extrapolating the
Tafel line to equilibrium potential provides the exchange current density (Io) of 2.0 x
10-6 Acm-2. The rate constant of the electron transfer (anodic oxidation in this case),
ko, is proportional to Io according to.
Io= nFAkoC (Equation 2. 8)
We obtained a rate constant of ko = 1.4 x10-3 cm.s-1 from the exchange current density
using equation 2.8. Our results show that RDE technique can be successfully applied
to highly concentrated electrolyte solutions. The data revealed defined limiting
currents from which the kinetics of the system can be deciphered.
2.4 Conclusions
A detailed study of the kinetics of the oxidation of ferrocene in a concentrated
lithium ion conducting electrolyte was carried out using cyclic and rotating disc
electrode voltammetry. The results obtained show that the ferrocene-ferrocenium
redox couple is reversible in this medium. The values for ferrocene and ferrocenium
ion diffusion coefficients were determined from these data. In addition, the electron
transfer rate constant (ko) and the exchange current density (Io) for the oxidation of
ferrocene were calculated. A comparison of the kinetic data obtained from the two
electrochemical techniques appears to show that the data from the RDE experiments
67
are perhaps more reliable, because they are collected under strict mass transport
control. A Tafel slope of c.a. 79mV/decade and a transfer coefficient α of 0.3
obtained from analysis of the RDE data suggest that the structure of activated
complex in the oxidation reaction of ferrocene is closer to that of the oxidized specie,
probably due to strong interactions with PF6- and carbonate solvents. Strong
interactions between the ferrocenium ion and the carbonate solvent is consistent with
the highly dipolar nature of the organic carbonates.
Our results indicate that useful electrochemical kinetic data for soluble redox
species in highly concentrated electrolyte solutions relevant to Li-ion batteries can be
obtained using the complementary CV and RDE techniques. Such kinetic data are
relevant to the studies of redox reagents for overcharge protection of Li-ion batteries,
particularly in simulation studies aimed at understanding their performance in
practical batteries, and in the development of improved materials.
68
2.5 References
(1) Gritzner, G. Pure Appl. Chem. 1984, 56, 461-466. (2) Page, J. A.; Wilkinson, G. J. Am. Chem. Soc. 1952, 74, 6149-6150. (3) Kealy, T. J.; Paulson, P. L. Nature 1951, 168, 1039-1040. (4) Tsierkezos, N. J.Solution. Chem. 2007, 36, 289-302. (5) Gennett, T.; Milner, D. F.; Weaver, M. J. J. Phys. Chem. 1985, 89,
2787- 2794. (6) Feng, G.; Xiong, Y.; Wang, H.; Yang, Y. Electrochim. Acta 2008, 53,
8253-8257. (7) Opallo, M.; Kukulka-Walkiewicz, J. Electrochim. Acta 2001, 46,
4235- 4242. (8) Reiter, J.; Vondrák, J.; Micka, Z. Electrochim. Acta 2005, 50, 4469-4476. (9) Cleary, J.; Bromberg, L. E.; Magner, E. Langmuir 2003, 19, 9162- 9172. (10) Abraham, K. M.; Alamgir, M. J. Electrochem. Soc 1990, 137, 1657- 1658. (11) Abraham, K. M.; Pasquariello, D. M.; Willstaedt, E. B.; ECS: 1990;
Vol. 137, p 1856-1857. (12) K.M.Abraham; D.M.Pasquariello; J.F.Rohan; C.C.Foo; US.Patent,
Ed.; EIC Laboratories, Inc.: USA, 1996; Vol. No 5858573. (13) Chen, J.; Buhrmester, C.; Dahn, J. R. Electrochem. Solid-State Lett.
2005, 8, A59-A62. (14) Behl, W. K.; Chin, D.-T. J. Electrochem. Soc 1988, 135, 21-25. (15) Wang, R. L.; Buhrmester, C.; Dahn, J. R. J. Electrochem. Soc 2006,
153, A445-A449. (16) Dahn, J. R.; Krause, L. J. J. Electrochem. Soc 2005, 152, A1283- A1289.
69
(17) Narayanan, S. R.; Bankston, C. P. J. Electrochem. Soc 1991, 138, 2224-2229.
(18) Behl, W. K.; Chin, D.-T. J. Electrochem. Soc 1988, 135, 16-21. (19) Golovin, M. N.; Woo, S. J. Electrochem. Soc 1992, 139, 5-10. (20) Abraham, K. M.; Jiang, Z. J. Electrochem. Soc 1996, 143, 1-5. (21) Richard, E. P.; Philip, J. E. J. Electrochem. Soc 1972, 119, 864-874. (22) Myland, J. C.; Oldham, K. B. J. Electroanal. Chem. 1983, 153, 43-54. (23) Shotwell, J. B.; Flowers, R. A. Electroanalysis 2000, 12, 223-226. (24) Murthi, V. S.; Urian, R. C.; Mukerjee, S. J. Phys. Chem. B 2004, 108,
11011-11023. (25) Petrocelli, J. V.; Paolucci, A. A. J. Electrochem. Soc 1951, 98, 291- 295. (26) Pournaghi-Azar, M. H.; Ojani, R. Electrochim. Acta 1994, 39, 953- 955. (27) Zhou, H.; Dong, S. Electrochimica Acta 1997, 42, 1801-1807.
70
Chapter 3
Elucidating the Mechanism of Oxygen Reduction for Lithium-Air Battery Applications
3.1. Introduction
The Lithium-air battery is one of the most energy dense, and environmentally
friendly, electrochemical power sources. Fully developed and optimally packaged Li-
air batteries could exceed specific energies of 2000 Wh/kg, versus a theoretical value
of 5200 Wh/kg, which is more than twice as much as any battery, primary or
secondary, presently known. The Li-air battery is composed of a Li metal anode and
an air cathode in which the cathode active material, oxygen, is accessed from the
environment. The first non-aqueous, rechargeable, Li-air battery1 used Li+-
conducting gel polymer electrolytes based on polyacrylonitrile (PAN) or
polyvinylidene fluoride (PVDF). In that battery Li2O2 was identified as a product of
the discharge reaction, which in presence of catalysts could be oxidized (recharged),
albeit at high overvoltages to oxygen and lithium metal. Later studies of Li-air
batteries utilized organic carbonate- and ether-based electrolytes of the types used in
Li metal and Li-ion batteries2. In a recent study Bruce and co-workers3 demonstrated
possibility of using Li2O2 as a positive electrode material in a Li/air battery which
was activated by initially charging (oxidizing) the peroxide to oxygen and lithium
metal. The electrochemical reduction of oxygen to superoxide and other oxides can
be taken advantage of practically as they can behave as Lewis bases, nucleophiles, as
a well as both oxidizing and reducing agents. These traits make the reduction of
71
oxygen desirable for energy production and storage. Previous electrochemical studies
of the oxygen reduction reaction (ORR) in organic solvents4-7 demonstrated that it is
possible to reduce molecular oxygen to superoxide (O2-) in a non-aqueous
environment. An identified distinction between the use of non-aqueous and aqueous
electrolytes is that in aqueous electrolytes the preferred reduction product is water or
hydrogen peroxide corresponding to a four or two-electron reduction of O2
respectively, as opposed to the formation of superoxide in organic electrolytes.
Almost all of the prior research in organic electrolytes utilized quaternary ammonium
cation (NR4+ where, R= ethyl, butyl etc.)-based salts as supporting electrolytes for ion
conduction. We are interested in understanding the electrochemistry of oxygen in
organic electrolytes in presence of alkali metal cations such as Li+, Na+ and K+ in an
effort to apply oxygen chemistry towards non-aqueous metal-air batteries, particularly
Li and Na batteries. These results together with the early investigations of oxygen
electrochemistry in non-aqueous electrolytes suggest that more than one product is
possible in the electrochemical reduction of non-aqueous Li-air batteries and that a
good understanding of the mechanism of oxygen reduction in organic electrolyte is
lacking. An in-dept study of the electrochemical redox behavior of O2, including the
kinetics and transport properties of the reduction and oxidation products in the
electrolyte, is important in further developing the Li-air battery. To this end, we have
initiated studies of the redox reactions of oxygen in non-aqueous electrolytes with the
objective of elucidating the roles of ion conducting salts and organic solvents on the
mechanisms of the corresponding reactions. We present a full account of our work in
acetonitrile. This solvent is not practically useful in a Li-air battery because its reacts
72
with Li metal. Despite this we chose it for this initial study because of previous
electrochemical studies of oxygen in this solvent and because of some initial
surprising results we obtained when a Li salt was used as the conducting salt. Most
of the early electrochemistry of molecular oxygen in organic solvents such as
dimethylsulfoxide(DMSO), dimethylformamide(DMF) and acetonitrile4-6, utilized
tetra alkyl ammonium perchlorate (NR4+ClO4) as the ion conducting salts leading to
similar overall results. We show here that there are significant differences in the
reduction mechanism and products when alkali metal salts are used. (Our results in
other practically more relevant organic electrolytes for the Li-air battery will be
published in the future). Using cyclic voltammetry (CV) and rotating disc electrode
(RDE) voltammetry we first studied O2 reduction in acetonitrile electrolyte solutions
containing both TBAClO4 and TBAPF6 to assess the influence of anion on oxygen
reduction. We then studied oxygen redox reactions in hexafluorophosphate-based
electrolytes of the formula A+ PF6- where A = TBA, K, Na, Li. We have discovered
that the electrochemistry of oxygen is strongly influenced by the nature of the cation
and very little by the anion in the conducting salt. Also our RDE studies reported
here represent the first application of this technique to elucidate the mechanism of
oxygen reduction in non-aqueous electrolytes and the results for the first time
provided detailed information on the influence of supporting electrolytes on the
kinetics and mechanisms of oxygen reduction in non-aqueous electrolytes.
73
3.2 Experimental
3.2.1 Chemical reagents.
All reagents were electrochemical grade unless stated otherwise stated. Battery grade
solvents and Lithium hexafluorophosphate (LiPF6) ((battery grade, >99.9%, H2O<
20ppm) were obtained from Ferro Corporation Cleveland, Ohio.
Tetrabutylammoniumhexaflourophosphate (TBAPF6), anhydrous acetonitrile
(MeCN), tetrabutylammonium perchlorate (TBAClO4), Potassium
hexafluorophosphate (KPF6), and Sodium hexafluorophosphate (NaPF6) were
purchased from Sigma-Aldrich, Allentown, PA.
3.2.2 Instrumentation.
The electrochemical experiments were performed with an Autolab (Ecochemie Inc.,
model-PGSTAT 30) potentiostat equipped with a bi-potentiostat interface in an
airtight electrochemical cell. The electrochemical cell designed and built in-house
consisted of a traditional 3-electrode system utilizing Ag/AgCl as the reference
electrode and platinum wire as the counter electrode. This reference electrode was
used instead of the Li foil electrode typically used in Li+ conducting electrolytes
because of its instability as a reference electrode in this electrolyte. The Ag/AgCl
gives a voltage of 2.93 V versus Li/Li+, as measured using a Li foil reference
electrode in a LiPF6 solution in organic carbonates. The cell also had inlet and outlet
valves for oxygen or argon purging. The cell was entirely airtight with exception of
the gas outlets, which were kept under pressure with the working gas. The glassy
carbon (3 mm diameter) working electrode employed for the cyclic voltammetry
experiments was polished with 0.5 and 0.05 mm alumina paste prior to the
74
experiments. For RDE experiments, the glassy carbon electrode was rotated with an
Autolab RDE rotor. All of the cyclic voltammetry experiments were initially
performed in an Ar-atmosphere glove box where H2O and O2 concentrations were
kept below 5ppm and temperature was held at 22 ± 2°C. For RDE experiments the
cell was brought outside of the glove box and placed in a glove bag purged with
Argon. The electrolyte solutions were first purged with argon, and the electrode was
cycled continuously until reproducible cyclic voltammetric profile was obtained. The
solutions were then purged with O2 for ORR measurements. All solutions were
prepared in the glove box. Conductivity measurements of all samples were carried
out using a 4-probe Thermo Orion conductivity cell from Thermo Fisher Scientific
Inc Waltham MA. Viscosity was measured using Ubbelohde viscometer purchased
from Technical Glass Products Inc NJ.
3.3 Results and Discussion
The roles of the TBA and alkali metal salts on the reduction properties of
molecular oxygen (O2) in acetonitrile were studied using cyclic (CV) and rotating
disk electrode (RDE) voltammetry. Cyclic voltammetry is a useful technique for
discerning kinetics, and mechanisms of electrochemical reactions. It is an
electrochemical potential sweep reversal method wherein a certain potential range is
swept at a known scan rate (measured in volt per second) in both the negative and
positive directions and the change in current is recorded. By applying appropriate
equations, the CV data can tell whether the reaction is nernstian (reversible), quasi-
reversible or irreversible. The RDE technique can be used in a complementary
fashion to discern the mechanistic details of the electrochemical processes. The same
75
disk electrode can be used to run both CV and RDE scans. The rotating disk,
hydrodynamic, technique utilizes convection as the mode of mass transport as
opposed to CV, which is governed by diffusion. Convection is a more efficient
means of mass transport with the result that the analytical data are more reproducible
and precise. In Table 3.1 we list the physical properties of the acetonitrile. In Table
3.2 the conductivities and viscosities of the electrolyte solutions used in this study are
presented.
Table 3.1: Physical properties of Acetonitrile
Some of these physical properties are used in calculating kinetic parameters
discussed below.
Table 3.2: Conductivity and Viscosity of the Electrolyte Solutions in acetonitrile
76
3.3.1 Oxygen Reduction in Tetrabutylammonium
hexafluorophosphate (TBA+PF-6)-Based Electrolytes
The full-range cyclic voltammograms (CV) scanned from -3V to 1V, for the
reduction of oxygen in 0.1M TBAPF6 and TBAClO4/MeCN are presented in fig 3.1a.
Potential (V)-3 -2 -1 0 1
Am
ps/c
m2
-4e-3
-3e-3
-2e-3
-1e-3
0
1e-3
2e-3
0.1M TBAClO4TBAClO 4 Background
0.1M TBAPF6
-2.0 -1.5 -1.0 -0.5 0.0 0.5
-4e-3
-3e-3
-2e-3
-1e-3
0
1e-3
2e-3
0.1M TBAPF60.1M TBAClO4
A
B
Ep1
Ep1
Ep2
Ep3
Ep3
Ep4
Figure 3.1: A) iR corrected voltammograms for the reduction of oxygen in 0.1M TBAPF6 (Black), 0.1M TBAClO4 (Blue) and the argon background (dotted) in MeCN. B) CVs in the –2 to +0.5 V range. All scans used a glassy carbon working electrode. Scan rate of 100mV/s.
77
CVs were first run under an inert atmosphere of argon to provide a
background voltammogram. As shown no appreciable current was observed under
argon over the full potential range of which oxygen redox reactions were investigated
confirming that the electrolyte contained no other electroactive species. The
similarity of the voltammograms is quite evident, the first reduction peak (Ep1) at c.a
(-0.9 V) is present in both CV’s with little difference in peak position. Polarizing the
electrode to further negative potentials a second reduction peak (Ep2) emerges at c.a
(–2.2V) also present in both electrolytes. The oxidation peak (Ep3) peak is very
similar in shape and size to that of Ep1 and only slightly separated in the ClO4-case.
This can only be attributed to the subsequent oxidation of Ep1 reduction products.
Ep4 is separated from Ep2 by almost 2V highlighting an irreversible reaction. In fact
Ep2’s reduction products are oxidized only at these high overpotentials. By
integrating the area under each peak we find the charge area. The charge area under
Ep1 and Ep3 peaks are similar. However expanding the electrochemical window to
encompass Ep2 we see a distinct loss in charge area under the anodic peak (Ep3).
Comparing this loss of charge to the area under Ep2 we find this area is proportional
to that of the loss (Table 3.3).
Table 3.3: Electrochemical Charge area under the peaks. Scan rate 100mV/s. Error ± 0.002C.
Thus implying a portion of the first reduction product formed on the electrode
undergoes a secondary irreversible reduction. The peak Ep4 appears if the second
Charge area (Coulombs x 10-3) Ep1 Ep2 Ep3 Ep1 - IR Ep2 -IR ClO4
- 4.98 1.45 3.47 4.91 4.30 PF6
- 4.72 1.31 3.30 3.7 3.3
78
reduction peak Ep2 is formed and hence we associate it with the oxidation of the
material generated at the electrode during this process. This can be clearly discerned
from CV in figure 3.1b, where the scan region is restricted to avoid Ep2. Generally
speaking these voltammograms are identical except for peak positions, which maybe
attributed to ohmic losses. This weak but noticeable oxygen reduction dependence on
the counter ion is evident by these shifts. ORR in perchlorate solution is slightly
positive by a 100mV indicating that oxygen reduction in the presence of
hexafluorophosphate is to some extent slightly polarized. This may be due in part to
the less coordinating nature of the PF6-, allowing the larger tetra butyl ammonium ion
to interact with dissolved oxygen. Generally the electrolyte/electrode interface is
affected by the nature of the counter ion. A summary of voltammetric results is
provided in Table 3.4.
Table 3.4: Voltammetric properties of O2/O2- redox couple in 0.1MTBAPF6 & TBAClO4/MeCN
These results indicate that the solvent/salt interactions are well coordinated in the case
of TBAClO4 leading to a structured double layer region, which is coupled to the ion
diffusion coefficients. We found the charge area ratio ( )ca QQ / under the peaks to
79
be over 89 ± 0.02 % for the CVs portrayed in Figure 3.1b. The peak potential
separation ∆Ep between the anodic and cathodic peak potentials for ClO4- and PF6
- are
presented in Table 3.4. These values are close to the theoretical value of 59mV for a
one-electron reaction. For a reversible process the peak width is given by the
following relationship.
p / 2 pR T
E E 2 .2n F
− =
Equation 3.1
Where Ep/2 is the half-peak potential at the half value of the peak current ip is the peak
potential, F is the Faraday constant and n is the number of electrons in the reaction.
Analysis of the CVs over the whole sweep ranges gave n values close to unity (see
Table 3.4) indicating that the number of electrons transferred in the reaction is one.
Possible reasons ∆Ep is slightly larger than the theoretical value are sluggish kinetics
due to Ohmic (iR) contributions (Etrue = Eactual –iR) at high scan rates. The magnitude
of the current (I) in a cyclic voltammetry is a function of temperature, concentration
C, electrode area A, the number of electrons transferred n, the diffusion coefficient
D, and the speed at which the potential is scanned V, all related by the Randles-
Sevcik equation (Eq 2).
5 3 2 1 2 1 2paI =(2.69×10 )n AD V C Equation 3.2
80
-1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4-1.5e-3
-1.0e-3
-5.0e-4
0.0
5.0e-4
25mVs-1
100mVs-1
300mVs-1
Potential (V)
Am
ps/c
m2
0.1 0.2 0.3 0.4 0.5 0.6 0.7-4e-3
-2e-3
0
2e-3
4e-3
TBAClO 4TBAPF6
n=1 n=2
v1/2 (V1/2/s1/2)
B
A
Ep1
Ep3
Figure 3.2: (A) Cyclic Voltammograms for the reduction of oxygen saturated 0.1M TBAPF6 /MeCN on GC electrode at sweep rates 0.1V/s (solid), 0.1V/s (long dash) and 0.025V/s (short dash), (B) Randles-Sevcik plot of peak current vs. square root of the scan rate for the curves in 0.1 M TBAPF6 & 0.1 M TBAClO4/MeCN.
81
Figure 3.2a displays the cyclic voltammograms for the reduction of oxygen
saturated TBAPF6/MeCN at different sweep rates. The reduction is reversible at all
sweep rates and there is only a slight shift in the peak position. The Randles-Sevcik
plots presented in Fig. 3.2b are linear and pass through the origin as per theory,
indicating a fast, diffusion controlled electrochemical process. The theoretical plots
of n =1 in figure 3.2b parallels the one-electron experimental plot implying that n=1
and that the first reduction involves the formation of superoxide (O2-). The presence
of superoxide in solution was confirmed qualitatively by adding Nitrotetrazolium
Blue Chloride tablet, which produced the characteristic purple color.
Figure 3.3(a) shows the typical steady-state voltammograms for O2 reduction
on a RDE in oxygen saturated 0.1 M TBAPF6 solution at various rotation rates. This
figure demonstrates that the current generated by this hydrodynamic method is much
larger than that generated in the CV under diffusion control. The much larger current
obtained using RDE reflects the efficiency of this method. We can easily determine
the limiting current, ilim, from these voltammograms. Also notice in the figure that
there is significant increase in cathodic current (i.e. O2 to O2-) while the amount of
anodic current (i.e. O2- to O2) is negligible essentially making the voltammogram
cathodic. This is due to the vast difference in the concentrations of O2 and the O2-
ions. The bulk solution contains O2, which is constantly supplied to the rotating
electrode while the superoxide ion’s concentration at the electrode is so minuscule
that little anodic current is produced. The Levich equation (3.3) establishes
relationship between current at the RDE and concentration of the analyte.
82
2 4 6 8 10 12 14 16 18 20 22-0.04
-0.03
-0.02
-0.01
0.00
TBAClO 4TBAPF6n=1n=2
-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5-0.010
-0.008
-0.006
-0.004
-0.002
0.000
0.002
400
900
1600250036004900
Potential (V)
w1/2
A
BA
mps
/cm
2
Figure 3.3: (A) Disk currents obtained in 0.1 M TBAPF6 MeCN during ORR in the anodic sweep at room temperature by various rotation rates at 100mV/s. (B) Levich plot of limiting current vs. square root of rotation in 0.1 M TBAPF6 & 0.1 M TBAClO4 in MeCN vs. Ag/AgCl at scan rate =100mVs-1.
83
In the Levich equation
( ) 2/3 1 /2 -1 /6limi = 0 .620 nFA D ω v C Equation 3.3
i lim is the limiting current density (A cm-2), n is the number of electrons involved in
the reaction, F is the Faraday constant (96,500 C mol-1), D is the diffusion coefficient
of oxygen in the solution, v is the kinematic viscosity of the solution (4.4 x 10-3cm2s-
1), C is the concentration of oxygen in solution (8.1mM)8,9 and ω is the angular
frequency (2
60
fπ). The RDE data provide insight into the number of electrons
transferred to the analyte by comparing the limiting currents to the rotation rate of the
electrode. Figure 3.3b displays the Levich plot for the reduction of oxygen from the
RDE data presented in figure 3a. A linear Levich plot passing through the origin
indicates that mass transfer of oxygen from the bulk solution to the electrode surface
controls the limiting current. The experimental Levich parallels the theoretical line
when n =1, where n is the number of electrons, indicating that the reduction of
oxygen at this electrode is a one electron process to form superoxide. These CV and
the RDE data are consistent with the reaction Scheme 1 for the reduction of O2 in
acetonitrile.
Scheme 3.1
Step1. O2 + TBA+ + e- = TBAO2
Step2. TBAO2 + TBA+ + e- = TBA2O2
84
The peak Ep1 in the CV corresponds to step 1 and Ep2 to step2. We calculated the
diffusion coefficient of O2 from the dependency of Ipc on 1/ 2V (from the Randles-
Sevcik equation). The diffusion coefficient for O2 (DO2) is found to be 2.2x 10-5 cm2
sec -1 in 0.1M TBAClO4 and 2.1x 10-5 cm2 sec -1 in 0.1M TBAPF6. These values are
very close to the previously reported values of 4.87x 10-5 cm2 sec -1 in 0.9 M TEABF4
and 2.07 x 10-5 cm2 sec -1 in acetonitrile containing 0.1M TEAP10,11. We also
calculated the diffusion coefficients of oxygen using the Levich method for ClO4- (2.3
x 10-5 cm2 sec -1) and PF6- (2.1 x 10-5 cm2 sec-1). Again the small difference in the
diffusion coefficient values between both the Randles–Sevcik and the Levich
equations maybe ascribed to the fact that the Randles-Sevcik equation does not
contain the term for mass transport control. Deviation from linearity at the lower
rotation rates in figure 3a is attributed to poor mass transport or slow kinetics. The
data presented above indicate that the most likely pathway for oxygen reduction is by
an initial one-electron transfer to O2 to form O2-. We can utilize the Stokes-Einstein
equation to calculate the theoretical diffusion coefficient for O2 in this electrolyte and
account for the small differences in the diffusion coefficients. The relationship
between diffusion coefficient and solution viscosity is given by the Stokes Einstein
equation (3.4).
6
kTD
aπυ= (Equation 3.4)
Where a is the effective hydrodynamic radius of oxygen, k is the Boltzmann
constant, and T is the temperature and µ is the dynamic viscosity. The latter was
85
calculated from the aforementioned kinematic viscosity and the solution density and
was found to be 0.384 cP. Using Stokes-Einstein relationship we calculated
DO2 = 2.6 x 10-5 cm2 sec -1. The O2 hydrodynamic radius used for this calculation
was 2.16A12. Randles-Sevcik can also be applied to obtain the diffusion coefficient
of the superoxide (DO2-) generated, values obtained were 8.4x 10-6 cm2 sec -1 in 0.1M
TBAClO4 and 9.x 10-6 cm2 sec -1 for 0.1M TBAPF6, approximately an order of
magnitude lower than that of O2. The diffusion coefficients of O2 and O2- are of
particular interest to understand and model mass transport of these species in the Li-
air battery. We investigated the nature of the reduction further using the Tafel
equation, which relates the rate of electrochemical reaction to overpotential according
to
o1-αnF
logi=logi + ηRT
(Equation 3.5)
A plot of log i versus overpotential (η) should be linear, from which the transfer
coefficient α, and the exchange current density io can be determined. Figure 3.4
shows cathodic Tafel plots obtained after the measured current is corrected for mass
transport to give the kinetic current. The kinetic current is calculated from the
equation,
lim
k=lim
i . ii
i - i (Equation 3.6)
where ik is the kinetic current density, i is the measured current density during O2
reduction, and ilim is the diffusion limited current density.
86
Log
i k
Over Potential (ηηηη) V
-0.8 -0.6 -0.4 -0.2 0.0
-6
-5
-4
-3
-2
TBAPF6 TBAClO 4
Figure3.4: Tafel plots for ORR at room temperature on a glassy carbon electrode at 2500 rpm for cathodic sweep 0V to -1.0V. (OCP: TBAPF6 -0.25V & TBAClO4 -0.34 vs Ag/AgCl). The Tafel region is indicated in red.
The Tafel slope is consistent with a reversible one-electron reduction to
superoxide (step 1), as the slope is very close to 120mVdec-1. This indicates that step
1 is rate determining. The reversibility of this step is evident from the kinetic data
listed in Table 3.5.
Anions (η): ClO4- (η): PF6
-
Tafel slope (mV/dec) 115 111
Exchange Current Density (io) (Acm-2) 4.33 x10-5
4.44x10-5
Rate Constant (ko) (cm.s-1) 2.82 x10-4 2.89x10-4
α 0.45 0.52 Table 3.5: O2/O2- kinetic parameters in 0.1M TBAPF6 & TBAClO4/MeCN
87
The exchange current, log io is defined as the intersection of the Tafel line and the y-
axis (log Ik). The standard rate constant ko is calculated from io using equation.
I0= nFAkoC (Equation 3.7)
We established that the anion has very little effect on the mechanism of reduction in
this media. Both perchlorate and hexafluorophosphate solutions exhibit very similar
electrochemical behavior.
3.3.2 Oxygen Reduction in Alkali Metal-Hexafluorophosphate (X+PF-6)
-
Based Electrolytes
The electrochemical behavior of oxygen in the presence of alkali metal
cations differed from that observed in the TBA-based electrolytes. Figure 3.5a,
illustrates the considerable difference in electrochemistry, when the TBA cation is
substituted with alkali metal cations such as lithium (Li), sodium (Na) and potassium
(K). Reversibility or lack thereof is a major difference between the TBA based
electrolytes and the alkali solutions. Reversible systems correspond to a half-wave
potential E1/2 that is near the peak potential Ep. Figure 3.5a (inset) illustrates the
irreversible nature of these systems. The reduction wave is broadened by the sluggish
kinetics, leading to a displacement in potential between E1/2 and Ep. The relevant
voltammetric properties are listed in Table 3.6. Although the CVs appear relatively
mundane these systems are a lot more complex upon closer inspection. The cathodic
peak is shifted from -0.84V as in the case of the TBA cation to ca. -0.7V at the scan
rate of 25mV/s respectively. The peak shifts are possibly the result of the relative
88
Lewis acidities of the cations. Both sodium and lithium cations are recognized as
hard Lewis acids due to their small ionic radii Li+ (0.90 Ǻ) and Na+ (1.16 Ǻ) and low
oxidation states. Hard Lewis acids have high charge densities on their surface and
tend to form ionic bonds with hard bases such as superoxide. The appearance of a
second cathodic peak, which is characterized by the plateau region at -1.5V, was a
distinct feature of the LiPF6 case. Investigation of the plateau region was conducted
by varying the scan rate (Fig. 3.5b).
89
-1 0 1 2
-8e-5
-6e-5
-4e-5
-2e-5
0
2e-5
4e-5
6e-5
8e-5
25mV/s 75mV/s200mV/s 500mV/s
-3 -2 -1 0 1 2 3
-6e-5
-4e-5
-2e-5
0
2e-5
4e-5
NaPF6LiPF6KPF6
A
B A
mps
/Cm
2
Potential(V)
Ep1Ep2
Ep3
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-100
-80
-60
-40
-20
0
V1/2
(v/s)1/2
I p (m
A)
Figure 3.5: (A) Cyclic voltammograms of oxygen reduction in 0.1M LiPF6 (dashed line), 0.1M NaPF6 (Solid) in MeCN. Scan rate of 100mV/s (-3V to 3V vs. Ag/AgCl). (B) Oxygen reduction voltammograms in 0.1M LiPF6 /MeCN on GC electrode at various sweep rates.
90
Ep2 is associated with the successive reduction of lithium superoxide to
lithium peroxide through the reaction in Step 2. Note that the appearance of the peak
Ep2 corresponding to lithium superoxide reduction is scan rate dependent. The lack
of this feature at low scan rates reveals that the kinetics of this process is extremely
rapid. On the reverse sweep to positive potentials this peroxide reduction product is
oxidized at high overpotentials via the reaction in step 3 (Ep3=1.3V). This peak is
analogous to peroxide oxidation observed in TBA salt solutions. This oxidation peak
is absent in the sodium CV probably as a result of the decomposition of sodium
superoxide to sodium peroxide via reaction in step 2 (scheme 3). Lithium peroxide
decomposes slightly in a similar manner to sodium peroxide but not to the same
extent. This explains the absence of the peak corresponding to the reduction of LiO2
at slow scan rates. The electrochemical reduction of oxygen in these solutions is
irreversible. The cathodic peak current is directly proportional to the square root of
scan rate (inset 5b), indicating a fast diffusion controlled reaction. The initial
electrode processes can be described by similar reactions for O2 reduction in presence
of both Li+ and Na+ as depicted in Scheme 3.2 and Scheme 3.3, respectively
Scheme 2:
Step 1 (Ep1) O2 + Li+ + e- = LiO2 Eo = 3.0V(Li/Li+)
Step 2 2 LiO2 = Li2O2 + O2
Step 3 (Ep2) LiO2 + Li++ e-= Li2O2 Eo = 3.1V
Step 4 (Ep3) Li2O2 = O2 + 2Li+ + 2e-
91
Scheme 3:
Step 1 (Ep1) O2 + Na+ + e- = NaO2
Step 2 2 NaO2 = Na2O2 + O2
Oxygen is reduced to lithium superoxide via reaction Step 1(Ep1). Knowing the
thermodynamic quantity ∆G (Gibbs free energy) the cell potential may be obtained
from the equation,
oG nFE−∆ = (Equation 3.8)
The calculated Eo is presented in scheme 3.2 for the lithium case.
According to equation (9)13 the transfer coefficient maybe approximated from
the difference between the peak potential and the half wave peak potential see table
3.6. The low αn values, which are not in the typical region of 0.5, suggest sluggish
kinetics due to formation of a passive oxide layer on the surface of the electrode.
p p/21.857RT 47.7
E -E = = mVαn α
(Equation 3.9)
The rate constant may also be calculated if the diffusion coefficient is known.
According to Nicholson14,15 an irreversible cathodic reaction modeled through the
relationship between Ip and 1/ 2V is linear, and is described by equation (3.10).
5 1/2 1/2 1/2pI =(2.99 x 10 ACD V)n(nα) (Equation 3.10)
92
The diffusion coefficient of oxygen in these alkali metal-based salts can be estimated
using this equation along with the calculated αn values. In equation (10) Ip is the peak
current, A is the area of the electrode, C is the concentration of oxygen, and V is the
scan rate. A plot of 1/ 2V vs. Ip shown in Figure 3.6 contains both the experimental
plots using data collected and the simulated plots for n equal to 1. From these plots
the number of electrons involved in the first reduction process is determined to be
one.
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-1.6e-4
-1.4e-4
-1.2e-4
-1.0e-4
-8.0e-5
-6.0e-5
-4.0e-5
-2.0e-5
0.0
LithiumPotassiumSodium
I p (
Am
ps\c
m2 )
V1/2(v/s)1/2
Figure 3.6: Experimental and theoretical (n=1) √v vs. Ip plots for in 0.1M LiPF6, NaPF6,
and KPF6 in MeCN.
This confirms that the overall reduction of oxygen in these salts is a one-
electron process to form an alkali metal superoxide. The diffusion coefficients of the
alkali salts are presented in Table 3.6.
93
Table 3.6: Voltammetric properties of 0.1M Li, Na & KPF6 in oxygen saturated acetonitrile. Scan rate 25mV/s
These are an order of magnitude lower then their TBA counterparts. The
standard rate constants of these reactions are calculated from the y-intercepts in figure
3.6. The ko values show that O2 reduction kinetics in sodium salt solutions are a
compared to lithium and potassium based electrolytes. The irreversibility of these
systems obvious from the lack of oxidation peaks in the sodium data even at high
scans rates points to the chemical decomposition of the first reduction product. In
order to understand the system in further detail we examined oxygen reduction as a
function of concentration. Figure 3.7 shows voltammograms for 1M APF6 (A= Li +,
Na+ & 0.5M K+) in MeCN, scanned at 100mV/s. For the cases of sodium and lithium
results are similar to those in 0.1M solutions although there was a shift in the anodic
peak position in the lithium case.
Cation Ep1 (V)
Ep2
(V) Ep3 (V) Eo (V) αn ko
(cms-1) Diffusion coefficient (cm2/s)
Li + -0.71 -1.30 1.8 -0.580 0.225 8.10e-5 3.77e-7
Na+ -0.76 - - -0.730 0.190 6.97e-4 1.03e-6
K+ -0.78 - 1.2 -0.677 0.230 1.97 e-4 2.30e-7
94
-3 -2 -1 0 1 2 3
-2e-4
-2e-4
-1e-4
-5e-5
0
5e-5
1e-4
1M Lithium 1M Sodium 0.5M Potassium
Cur
rent
(Am
ps\c
m2 )
Potential (V) Figure 3.7: Cyclic Voltammograms for the reduction of oxygen saturated 1M XPF6 (X= Li+, Na+& 0.5M K+) in MeCN on GC electrode at 100mV/s.
Increasing the concentration of alkali salts facilitates oxidation of the
reduction products. Increased concentrations of cations stabilize the superoxide and
peroxide products. The electrochemistry of oxygen is influenced by the cation size,
increasing the cation size from lithium to potassium alters the cyclic voltammogram.
Potassium is a larger alkali metal (r =2.20 A). The CV in 0.1M KPF6 it is comparable
to Li and Na salt solutions. At high concentrations it is reminiscent of the TBA
electrolytes; notice two reduction peaks followed by two subsequent oxidation peaks
observed on the return sweep in the range. Applying RDE voltammetry to these
systems was unsuccessful. This is illustrated in Figure 3.8 where it is interesting to
note that there is little increase in current density as the electrode is rotated, in fact the
95
current decreases. It appears that the reduction product passivates the electrode and
as the rotation rate increases so does the passivation rate. The electrode appears to
passivate quicker than mass transport limit can be reached. The same behavior is
observed for NaPF6 although it occurs much faster. Thus the RDE data provide
additional support to the notion that the reduction products of oxygen passivate the
electrode. Figure 3.9 shows the typical steady-state voltammograms obtained on the
RDE in O2-saturated solution containing 0.1 & 1M KPF6. These voltammograms do
not show a clear limiting current due to mass transport limitations. However, the
formation of a slight horizontal current plateau during oxygen reduction is observed.
The formation of this plateau at potentials less negative than -1V is indicative of a
competitive secondary reduction at the electrode. This maybe the reason behind the
overlapping oxygen reduction peaks especially in the 1M solution.
96
Potential(V)
-1.5 -1.0 -0.5 0.0 0.5 1.0
Cur
rent
(A
mps
\cm
2 )
-3e-5
-2e-5
-1e-5
0
100rpm900rpm1600rpm 4900rpm
-2 -1 0 1
-6e-5
-4e-5
-2e-5
0
2e-5
100rpm900rpm1600rpm 3600rpm
0.1M NaPF6
0.1M LiPF6
Figure 3.8: Steady voltammograms for the reduction of oxygen in 0.1M LiPF6 & NaPF6 in MeCN at various rotation rates at 100mV/s.
97
-2 -1 0 1 2
-5e-4
-4e-4
-3e-4
-2e-4
-1e-4
0
100rpm 900rpm 1225rpm 2500rpm
Potential(V)
1M KPF6
-1 0 1 2
-4e-5
-3e-5
-2e-5
-1e-5
0
1e-5
100rpm 400rpm 900rpm 1600rpm
0.1M KPF6C
urre
nt(A
mps
)\cm
2
Figure 3.9: Disk currents obtained in 0.1 & 1M KPF6 MeCN during ORR in the anodic sweep at room temperature at various rotation rates. All scans used a glassy carbon working electrode at a scan rate of 100 mV
98
In summary, the data we obtained in TBAClO4 and TBAPF6 based
electrolytes reveals that the anion has little or no effect on the redox processes. In
TBA salt solutions the first reduction process is a one-electron reversible reduction of
oxygen to form the superoxide. The superoxide can be reduced to the peroxide
irreversibly at lower potentials. Alkali metal hexafluorophosphate APF6 (where A is
Li+, Na+ and K+) solutions were investigated to establish the effect of cations on
oxygen electrochemistry. By replacing the larger TBA+ with alkali metal cations the
reversible nature of oxygen reduction is severely suppressed. The reduction reaction
in solutions containing the smaller cations Li+ and Na + is irreversible. In LiPF6
solutions O2 is irreversibly reduced first by a one-electron process to form LiO2,
followed by a second one-electron reduction to Li2O2 which appears to passivate the
electrode surface making the reaction irreversible. It also appeared that the LiO2
formed on the electrode surface chemically decomposes to Li2O2. However, there is
a finite lifetime for the LiO2 on the electrode surface with the result that at high scan
rates the reduction of LiO2 to Li2O2 can be observed. Both LiO2 and Li2O2 can be
oxidized at high overvoltages to oxygen and lithium. Oxygen reduction in NaPF6 is
also a one-electron first step to form NaO2, which appears to passivate the surface as
well as decomposing rapidly to Na2O2 hence the complete lack of oxidation.
The sodium oxides cannot be oxidized even at high overvoltages, except in
highly concentrated electrolyte solutions. Potassium a slightly larger alkali metal
(radius =2.2 Ǻ) and displays a voltammogram that is somewhat quasi reversible as
verified by the increase of anodic current in comparison to lithium and sodium.
Although the oxygen reduction is not as reversible as that in the tetra alkyl
99
ammonium salt solutions the reduction of O2 to KO2 and KO2 to K2O2 is observed as
two distinct peak as opposed to the mix potential regions of lithium and sodium at 1m
concentration. Both of these oxides are electrochemically oxidized at significant
overpotentials.
These results maybe explained in terms of the charge density on the surfaces
of the cations. The smallest of the cations Li+ is a good Lewis acid capable of
forming a very strong ionic bond with the superoxide ion. Increasing the cation size
from Li+ to TBA+ (TBA+< K+<Na+<Li+) the positive charge density (charge per unit
volume) on the ion decreases, as does the relative Lewis acidity, leading to weaker
interactions with the superoxide ion. This has several consequences: The TBAO2 is
soluble in the electrolyte and the redox reaction is reversible. The KO2 formed
appears to have partial solubility in the electrolyte with the result that the redox
processes are somewhat reminiscent of that in the TBA solutions. The smaller Li and
Na cations are stronger Lewis acids and they form ionic bonds with oxides leading to
their precipitation on the electrode surfaces. This surface coverage of the electrode
by the O2 reduction products passivates the electrode, shuts down the reduction and
renders the reaction irreversible.
3.4 Conclusions
Our results show that the reduction and subsequent oxidation of O2 in
acetonitrile-based electrolytes is strongly influenced by the cation of the conducting
salt used. A practical outcome of the results from this work to the lithium-air battery
is that it would be advantageous to use a mixture of Li and K and/ or TBA salts as
supporting electrolytes in order to dissolve the oxygen reduction products. This in
100
turn would increase the amount of oxygen that can be reduced to deliver higher
capacity. Dissolving the reduction products would also promote reversibility of O2
reduction, which would increase the battery’s rechargeability. Our results also show
that useful electrochemical kinetic data for soluble redox species in highly
concentrated electrolyte solutions relevant to Li batteries can be obtained using the
complementary CV and RDE techniques. Such kinetic data are relevant to the studies
of the Li-air battery as well as others containing soluble electrode materials especially
for battery simulation studies aimed at understanding the performance of practical
batteries, and generally for the development of improved materials.
3.5 References (1) Abraham, K. M.; Jiang, Z. J. Electrochem. Soc 1996, 143, 1-5. (2) Read, J. J. Electrochem. Soc 2006, 153, A96-A100. (3) Ogasawara, T.; Debart, A.; Holzapfel, M.; Novak, P.; Bruce, P. G. J. Am. Chem. Soc. 2006, 128, 1390-1393. (4) Johnson, E. L.; Pool, K. H.; Hamm, R. E. Anal. Chem. 1966, 38, 183- 185. (5) Maricle, D. L.; Hodgson, W. G. Anal. Chem. 1965, 37, 1562-1565 . (6) Peover, M. E.; White, B. S. Electrochimica Acta 1966, 11, 1061-1067. (7) Sawyer, D. T.; Roberts, J. L. J. Electroanal. Chem. 1966, 12, 90-101. (8) Achord, J. M.; Hussey, C. L. Analytical Chemistry 1980, 52, 601-602. (9) Sawyer, D. T.; Chiericato, G.; Angelis, C. T.; Nanni, E. J.; Tsuchiya, T. Anal. Chem. 1982, 54, 1720-1724. (10) Kishioka, S.-y. Electroanalysis 2001, 13, 1161-1164. (11) Tsushima, M.; Tokuda, K.; Ohsaka, T. Anal. Chem. 1994, 66, 4551- 4556.
101
(12) Bader, R. F. W.; Henneker, W. H.; Cade, P. E. The Journal of Chemical Physics 1967, 46, 3341-3363. (13) A.J.Bard Electrochemical Methods Fundamentals and Applications; 2 ed.; John Wiley & Sons: New York, 2001; Vol. (14) Nicholson, R. S. Anal. Chem. 1965, 37, 1351-1355. (15) Nicholson, R. S.; Shain, I. Anal. Chem. 1964, 36, 706-723.
102
Chapter 4
Influence of Non-aqueous Solvents on the Electrochemistry
of Oxygen in the Rechargeable Lithium-Air battery
4.1 Introduction
The non-aqueous, rechargeable Li-air battery, introduced in 19961has emerged
as a major candidate for future alternative energy source. It is actively being
developed worldwide because of its potential to deliver ultrahigh energy density in a
battery that is low cost and environmentally friendly. In the first rechargeable Li-air
cell reported by Abraham1, composed of a Li metal anode, a polyacrylonitrile-based
gel polymer electrolyte2,3 and a porous carbon cathode, Li2O2 was identified as the
discharge product. The formation of Li2O2 is consistent with the open circuit voltage
(OCV) of about 2.9 V measured for the cell (1) and the theoretical voltages calculated
for possible Li-air cell reactions depicted in equations 1-3.
2Li + O2 = Li2O2; ∆Go = -145 kCal (Eo = 3.1 V) (1)
4Li + O2 = 2Li2O; ∆Go = -268 kCal (Eo = 2.91) (2)
Li + O2 = LiO2 ; ∆Go = -70 kcal (Eo= 3. 0V) (3)
Equations 1-3 reveal that two other products besides Li2O2 can be formed from the
reduction of oxygen. Recently, we have shown3 that the first product of the reduction
103
of oxygen in non-aqueous electrolytes is superoxide, O2-, involving a one-electron
process. We also found that the half-life of the superoxide depends on the nature of
the supporting electrolyte cation present in the electrolyte solution. In presence of
tetra butyl ammonium cations (Bu4N+) in acetonitrile solutions, the superoxide,
Bu4NO2, is extremely stable and resists further reduction to O22- or O2-. On the other
hand, in presence of Li+ ions, the superoxide, LiO2, is unstable with very short half-
life and decomposes to Li2O2 and O2. The LiO2 that survives decomposition can be
reduced to Li2O2.
The electrochemistry of O2 in presence of Na+ is somewhat similar to that in
presence of Li+, except that the NaO2 first formed appears to decompose very rapidly
to Na2O2. Recent data5 suggest that Li2O is probably formed in some Li/O2 cells
from the reduction of Li2O2. The rechargeable Lithium-air battery research is in its
infancy and a lot of further work remains to be done to fully elucidate the cell
chemistry involved in discharge/charge cycling, and to bring this technology to
practicability. A number of research groups have heeded the call and investigated
various aspects of this battery. The work so far can be divided into three major
categories; 1) Li-air cells with liquid and solid electrolytes, 2) porous electrode
materials and structures, and cell performance evaluation, and 3) catalysis of cell
reactions
Jeffery Read has contributed to liquid electrolytes6-8 for Li-air batteries.
Having conducted an exhaustive review of solvent properties he found electrolyte
formulation as having the largest influence on cell performance including the nature
of the reduction products. Discharge capacity is dependent on O2 solubility, which
104
led him to suggest ether-based electrolytes for improved cell performance. Abraham
et al2 studied low volatile organic liquid and polymer electrolytes for the Li-air
battery. Hydrophobic ionic liquids9,10 have been studied as electrolytes
demonstrating good lithium stability and high cell discharge capacities. Another
avenue of investigation involved applying existing electrolytes from conventional Li-
ion batteries to the Li-air11 battery. Recently, the usefulness of solid electrolytes for
Li-air batteries has been demonstrated with an all-solid-state rechargeable Li-air
battery12. Protected lithium electrodes (PLE) stabilized by lithium ion conductors13
have been applied successfully in both aqueous and non-aqueous Lithium batteries.
Finally, low loading of a very high surface area carbon on nickel foam14 has
demonstrated the highest discharge capacity thus far. Since the discharge products of
the Li-air battery are insoluble in most organic electrolytes, a porous electrode
structure with appropriate morphology, surface structure, pore volume and surface
area is crucial for the oxygen reduction reaction (ORR) and rechargeability of the Li-
air cell.
Abraham et al clearly established in their first paper1 that the Li/O2 cell is
rechargeable. They found that in the absence of a catalyst of pyrolyzed cobalt
phthalocyanine, (Co-Pc) the recharge occurs at about 4 V, with a large hysteresis
between charge and discharge voltages. The hysteresis was reduced and the
charge/discharge efficiency increased with the Co-Pc-based catalyst. Recent
investigations have employed manganese oxide (MnO2) catalysts15 although the
charge voltages in these cells are similar to the uncatalyzed cells. Our recent studies
have revealed that the Li/O2 cell can be recharged with high efficiency without a
105
catalyst using an appropriate porous carbon electrode3,16. Interestingly charge
voltages of these uncatalyzed cells are similar to those of the MnO2 catalyzed cells
with both of these cell exhibiting higher charge voltages than the cobalt-catalyzed
cells. Clearly, a full understanding of the mechanism of the cell discharge reaction
mechanism and rechargeability is still lacking.
In this chapter we report on the results of a detailed study of the influence of
non-aqueous solvents on O2 electrochemistry. Our results have shown a relationship
between the Lewis basicity of the solvents, measured by their Gütmann donor
numbers (DN)17, the Lewis acidity of the cations, and the relative stabilities of the
oxygen reduction products in presence of TBA+ and Li+, and their rechargeability.
These results complementing our recently published results3 on the influence of
supporting electrolyte cations on O2 reduction products are expected to provide the
ability to systematically design and select new electrolytes for the rechargeable Li-air
battery. The structural formulas of the four solvents studied and their acronyms used
here are:
Figure 4.1: Solvent Structures
106
4.2 Experimental
4.2.1 Materials
Anhydrous acetonitrile (MeCN), Dimethyl sulfoxide (DMSO), 1,2-Dimethoxyethane
(DME) and Pursis Tetraethylene glycol dimethyl ether (TEGDME) were purchased
from Sigma-Aldrich, Allentown, PA. All chemicals were dried with Lithium and
were stored and prepared in an MBraun dry box filled with purified argon where the
moisture and oxygen content was less than 1ppm. The dried solvents were stored
over 0.3 or 0.4nm molecular sieves and prior to actual measurements all solvents
were degassed under vacuum.
Tetrabutylammonium hexafluorophosphate (TBAPF6) electrochemical grade,
≥99.0% (Fluka, puriss grade) from Sigma-Aldrich, Allentown, PA was dried under
reduced pressure at room temperature. Lithium hexafluorophosphate (LiPF6) (battery
grade, >99.9%, H2O< 20ppm) was obtained from Ferro Corporation Cleveland, Ohio.
4.2.2 Electrochemical Experiments
The electrochemical experiments were performed with an Autolab
(Ecochemie Inc., model-PGSTAT 30) potentiostat equipped with a bi-potentiostat
interface in an airtight electrochemical cell. The electrochemical cell designed and
built in-house consisted of a traditional 3-electrode system utilizing Platinum (Pt)
mesh as the reference electrode and Pt mesh as the counter electrode. This reference
electrode was used because of the instability of Li foil typically used in Li+
conducting electrolytes as a reference electrode because of its reaction in acetonitrile.
The Pt reference electrode provided stable potentials and was calibrated with
107
reference to the ferrocenium ion/ferrocene couple (Fc+/Fc) in each electrolyte studied,
which in turn was calibrated to Li/Li+ in a stable ethylene carbonate/dimethyl
carbonate-based electrolyte. The cell also had inlet and outlet valves for oxygen or
argon purging. The cell was entirely airtight with exception of the gas outlets, which
were kept under pressure with the working gas. The glassy carbon (5 mm diameter)
working electrode employed for the cyclic voltammetry experiments was polished
with 0.5 and 0.05 mm alumina paste prior to the experiments. For RDE experiments,
the glassy carbon electrode was rotated with a Pine AFCPRB RDE rotor. All of the
cyclic voltammetry experiments were initially performed in an argon-atmosphere
glove box where H2O and O2 concentrations were kept below 1ppm and temperature
was held at 22 ± 2°C. For RDE experiments the cell was brought outside of the glove
box and placed in a glove bag purged with Argon.
The electrolyte solutions were first purged with argon, and the electrode was
cycled continuously until reproducible cyclic voltammetric profile was obtained. The
solutions were then purged with O2 for ORR measurements. The electrochemical
impedance measurements were performed with the Autolab PG 30 supplied with a
FRA 2 module for impedance measurements. The impedance spectra were measured
in the frequency range from 100 mHz to 100 kHz at open circuit potential with an AC
voltage amplitude of 5mV. Conductivity measurements of all samples were carried
out using a 4-probe Thermo Orion conductivity cell from Thermo Fisher Scientific
Inc Waltham MA. Conductivity data for the solutions of 0.1M NBu4PF6 and LiPF6 in
dimethyl sulfoxide (DMSO), acetonitrile (MeCN), 1,2 Dimethoxyethane (DME),
108
tetraethylene glycol dimethyl ether (TEGDME) are summarized in Table 4.1. All
measurements were carried out at room temperature (22±2).
Table 4.1: Conductivity of the Electrolyte Solutions
4.3 Results and Discussion
4.3.1 ORR in Selected Non-Aqueous Electrolytes.
Electrolytes based on aprotic non-aqueous solvents are the ideal medium to
investigate the oxygen reduction reactions (ORR) relevant to the Li-air battery. An
environment free of protons could enable the full reduction of oxygen, essential to
realize the full energy density of the Li-air cell without interference from protonated
intermediates or products. Our previous work3 revealed that the three possible O2
reduction products in the Li-air battery, LiO2, Li2O2 and Li2O are highly polar.
Therefore, appropriate polar solvents are required to dissolve these products in order
to avoid their precipitation and passivation of the electrode surface. However, there
is no metric currently existing to select the optimum non-aqueous solvent for the
rechargeable Li-air battery.
109
a Goldfarb et al., Ref.28 b Rivas et al., Ref.29 c Lago et al., Ref.30 d Chemistry of Non aqueous Solutions., Ref.31 e Aminabhavi et al.,Ref.321 f Marcus Properties of Solvents.,Ref.33 g Sawyer et al.,Ref.18 h Read,.Ref.5
Polar solvents such as sulfoxides (R2S=O), ethers (R–O–R) and nitriles
(RC≡N) are potentially useful candidates as they may dissolve O2 reduction products
at least partially to promote rechargeability, but there is no guiding principle presently
available to select the best solvent or family of solvents Table 4.2 lists the four
solvents with widely varying properties, particularly donor numbers (DN) that are a
measure of solvent basicity, investigated in this work. We have purposely chosen
these solvents with the goal of identifying a fundamental property or properties that
can be used as the metric to select solvents with optimum properties for the Li-air
battery.
110
4.3.2 ORR in TBAPF6 solutions in DMSO, DME and MeCN
Dimethyl sulfoxide (DMSO) is a highly polar versatile solvent, which displays high
salt solubility to produce well-conducting solutions with a wide electrochemical
window (Figure 4.2A). This figure also displays a cyclic voltammogram (CV) for the
reduction of oxygen in a 0.1M TBAPF6/DMSO electrolyte. The peak potential
separation ∆Ep between the anodic (Epa = 2.40V) and cathodic (Epc = 2.34 V) peaks is
60mV and the charge area ratio ( )ca QQ /
under the peaks is close to unity. These
results indicate that O2 reduction in the presence of TBA+ ions is reversible.
111
-3e-3
-2e-3
-1e-3
0
1e-3
2e-3
MeCN DME
B
Potential (V) vs. Li/Li+
1 2 3 4
Cur
rent
(A
mps
/Cm
2 )
-8e-4
-6e-4
-4e-4
-2e-4
0
2e-4
4e-4
6e-4
8e-4
ArgonTBA+
Epa
Epc
A
Figure 4.2: A) Cyclic voltammograms for the reduction of oxygen in 0.1M TBAPF6 (Red, iR corrected) and the argon background (dotted) in DMSO. B) Cyclic voltammograms (iR un-corrected) for the reduction of oxygen in 0.1M TBAPF6/MeCN (Black), DME (Blue). Scan rate 100mV/s.
112
Figure 4.3 portrays a Randles-Sevcik (RS) plot of this ORR. The Randles-Sevcik
equation (1) describes the relationship between the current and scan rate of a
reversible electrochemical reaction. The magnitude of the current (I) is a function of
temperature, T, the oxygen concentration in solution, C, (2.1 mM)18, electrode area A,
the number of electrons transferred n, the diffusion coefficient D, and the rate, V, at
which the potential is scanned ( scan rate).
5 3 2 1 2 1 2
paI =(2.69×10 )n AD V C (Equation 4.1)
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-2.5e-4
-2.0e-4
-1.5e-4
-1.0e-4
-5.0e-5
0.0
Randles Sevcik Plot TBA+
n=1 Randles Sevcik Plot
Cur
rent
I (A
mps
)
Scan Rate (V/s)
Figure 4.3: Randles-Sevcik plot of peak current vs. square root of the scan rate in 0.1 M TBAPF6 /DMSO.
The plot of experimental data versus a theoretical Randles-Sevcik plot shows that it is
in close agreement with the n =1 theoretical plot, thereby indicating that Epc is a one-
113
electron reduction process. The plot linearity also suggests that this is a mass
transport limited process. This behavior is identical to that we previously found in
TBAPF6/ acetontrile3 and by others in TBAClO4 solutions18. The reduction of O2 in
DME/TBAPF6 and MeCN/TBAPF6 exhibits similar behavior as shown in figure 4.2B
indicating the general nature of the mechanism of O2 reduction in TBA+-containing
solutions. The O2 reduction potential and the associated current varied slightly in the
different electrolytes probably due to the different O2 solubilities and reduction
kinetics. The voltammograms obtained from the RDE experiments were analyzed
using the Levich equation (2) which defines the relationship between current at a
rotating disk electrode RDE and the angular frequency (ω) of rotation of the
electrode.
( ) 2/3 1 /2 -1 /6limi = 0 .620 nFA D ω v C (Equation 4.2)
In equation 4.2 ilim is the limiting current density (amps), n is the number of electrons
involved in the reaction, F is the Faraday constant (96,500 C mol-1), D is the diffusion
coefficient of oxygen in the solution, v is the kinematic viscosity of the solution (1.9
x 10-3cm2s-1)19. In RDE voltammetry, steady state is reached quickly eliminating
double layer charging. Also mass transfer affects are eliminated, as mass transfer
rates are much larger than diffusion rates allowing for accurate kinetics calculations.
Figure 4.4 displays the Levich plot for the reduction of oxygen in 0.1M
TBAPF6/DMSO; its linearity indicates that mass transfer of oxygen from the bulk
solution to the electrode surface controls the limiting current. The experimental
Levich plot parallels the theoretical line
114
A
5 10 15 20 25
IL
(Am
ps)
-8e-4
-7e-4
-6e-4
-5e-4
-4e-4
-3e-4
-2e-4
n =1Experimental
ω
-0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00
log I
k
-7
-6
-5
-4
-3
-2
120mV/Dec (n=1)
η
Figure 4.4: Levich plot of limiting current vs. square root of rotation in 0.1 M TBAPF6/DMSO scan rate=100mVs-1 (Inset Tafel plot).
when n =1, which is consistent with the CV data. The kinetic nature of the reaction
can be further investigated using the Tafel equation,
k o1-αnF
logi =logi + ηRT
(Equation 4.3)
A plot of log ik versus overpotential (η) should be linear, from which the transfer
coefficient α, and the exchange current density io can be determined. The inset in
figure 4.4 shows cathodic Tafel plots obtained after the measured current is corrected
for mass transport to give the kinetic current. The kinetic current is calculated from
the equation,
lim
k=lim
i . ii
i - i (Equation 4.4)
115
where ik is the kinetic current density, i is the measured current density during O2
reduction, and ilim is the diffusion limited current density from the Levich plot. The
Tafel slope is consistent with a reversible one-electron reduction to superoxide, as the
slope is very close to 120mVdec-1. This indicates that Epc is the rate-determining step
(rds). The reversibility of this step is evident from the kinetic data listed in Table 4.5.
The kinetic current density, ik, the diffusion-limited current ilim density, and
the measured current density, i, are related through the Koutecky-Levich equation
2/3 1/2 1/6lim
1 1 1 1 1
0.62K K o oi i i i nFAD v Cω −= + = + (Equation 4.5)
The inverse kinetic current density, 1/ik, can be obtained from the intercept of
Koutecky-Levich plot Fig 4.5. Reasonably linear plots are obtained (see the insets) at
all measured potentials where ORR is expected to be under the mixed
kinetics/diffusion control, and the linear plot under the pure diffusion control
intercepts close to zero. Determination of ik at different values of E allows
determination of the standard rate constant k° at different potentials where the rate of
electron transfer is sufficiently slow (equilibrium) to act as a limiting factor and when
the electron transfer is rapid in the limiting-current region. Standard rate constants
varied from 3.8x10-2 to 4 x10-3cm-1 and 3x10-3 to 6 x10-4cm-1 for DMSO and MeCN
respectively.
116
1.5 2.0 2.5 3.0 3.5 4.0
i-1
-0.020
-0.015
-0.010
-0.005
0.000
0.005
400 900 1600 2500 3600
0.04 0.06 0.08 0.10 0.12 0.14 0.16-3e+3
-3e+3
-2e+3
-2e+3
-1e+3
-5e+2
0
-2.46V-2.36V -2.27V -2.17.V
Potential vs Li/Li+
Am
ps/c
m2
1
ω−
1.0 1.5 2.0 2.5 3.0 3.5 4.0-0.05
-0.04
-0.03
-0.02
-0.01
0.00
0.01B
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18-3.5e+4
-3.0e+4
-2.5e+4
-2.0e+4
-1.5e+4
-1.0e+4
-5.0e+3
0.0
-2.27V-2.36V-2.46V -2.50V
i-1
1
ω−
i-1
A
Figure 4.5. Current-voltage curves measured at 100 mV/s on a GC rotating disk electrode (400-3600rpm) for oxygen reduction in (A) 0.1M TBAPF6/DMSO (B) 0.1M TBAPF6/MeCN. Insets: Koutecky- Levich plot at different potentials in kinetic-diffusion region of the polarization curve.
117
We can describe the ORR mechanism in TBAPF6 solutions according to the reactions
in Scheme 1, involving a
Scheme 4.1:
Cathodic (Epc). O2 + TBA+ + e- = TBAO2 (6)
Anodic (Epa). TBAO2 - e- = TBA+ +O2 (7)
one electron reduction of oxygen to superoxide (O2-) and subsequent reoxidation of
superoxide to oxygen. An explanation for the reversible O2 reduction process in TBA
salt solutions and the superior stability of the superoxide, O2-, in presence of TBA+ in
the various solvents is presented later in this chapter.
4.3.3 ORR in LiPF6 solutions in DMSO, DME, MeCN, and TEGDME.
The ORR results obtained in these electrolytes will show clearly that the O2 reduction
mechanism in Li+-containing electrolytes is different from that seen in presence of
TBA+. In addition, these results will demonstrate the subtle influence of the solvent
on the mechanistic details of the O2 reduction reactions in Li+-containing electrolyte
solutions as well as the rechargeability of the reduction products. We have found that
the voltammetric data in DMSO is especially instructive to unambiguously map the
O2 reduction mechanism in Li+-containing organic electrolytes relevant to the
rechargeable Li-air battery.
118
Figure 4.6 illustrates O2 reduction in 0.1M LiPF6/DMSO. This figure comprises four
separate CV’s overlaid. Each CV corresponds to a defined electrochemical window
over which the voltammogram was scanned.
Figure 4.6: Cyclic voltammograms (iR corrected) for the reduction of oxygen in 0.1M LiPF6/DMSO at various potential windows. All scans used a glassy carbon working electrode. Scan rate of 100mV/s.
The shortest window is shown in dark yellow (2.57- 4.5 V) in which the scan was
reversed at the half-peak potential Epc1/2 (2.57V) of the first cathodic peak in order to
examine the associated anodic features. Reversing the sweep at Epc1/2 resulted in two
clear anodic peaks, Epa1 at 2.75V followed by a broad peak (Epa2) at 3V. Expanding
the cathodic scan to the peak potential Epc1 (2.45V) produces an increase of the
current in the following anodic Epa1 & Epa2 peaks becoming similar in magnitude.
119
Two anodic peaks resulting from a single cathodic peak suggests a dual step
reduction mechanism from the very beginning. A one-electron reversible process is
characterized by 56mV difference between the cathodic peak and half-peak potential.
For this system the potentials (|Epc1 - Epc1/2|) are separated by 100mV, demonstrating
the complexity of this process. Upon scanning cathodically further, the current slope
changes at 2.12V, Epc2 (blue), signifying another electrochemical event. Reversing
the scan subsequently in the positive direction results in the disappearance of Epa1 and
increase in Epa2 peak current. This suggests that the first reduction product is
consumed and converted to the second reduction product, which is oxidized at Epa2.
Finally the cathodic sweep was allowed to continue towards 1.35V (Red line)
where it was reversed. The corresponding anodic scan consists of two broad
overlapping peaks. Similar to the blue scan Epa1 is absent and the magnitude of Epa2
decreased. The new anodic peak Epa3 that emerged is believed to be due to the
oxidation of the product formed from the reduction at Epc2.
As these reactions are irreversible Randles Sevcik & Levich treatments cannot
be applied to these CV data. We have deconvoluted the data using the Nicholson &
Shain relatioship20 (equation (8)) developed for irreversible electrochemical reactions,
5 1/2 1/2 1/2
pI =(2.99 x 10 ACD V)n(nα) (Equation 4.8)
The symbols in equation 4.8 have their usual meaning. Figure 4.7a clearly shows that
the number (n) of electrons transferred in the first reduction reaction is one since the
theoretical n=1, plot follows the experimental data.
120
B
-4.8 -4.7 -4.6 -4.5 -4.4 -4.3 -4.22.46
2.48
2.50
2.52
2.54
2.56
2.58
120 mV/dec220 mV/dec
log i
Pot
entia
l (V
) vs
. Li/L
i+
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-3e-4
-3e-4
-2e-4
-2e-4
-1e-4
-5e-5
0
Experiementaln=1
Cur
rent
I (A
mps
)
Scan Rate (V/s)
A
B
Figure 4.7: (A) Peak current vs. square root of the scan rate in 0.1 M LiPF6/DMSO. (B) Cathodic Tafel plot obtained in 0.1 M LiPF6/DMSO during ORR. Scan rate = 10mV/s.
121
The best theoretical fit was obtained using a transfer coefficient α =0.5 which
is a typical value for reversible reactions. This suggests that the first one-electron
reduction of O2 in DMSO/LiPF6 is substantially reversible. Tafel analysis can also be
used to obtain further insight. Tafel plots for ORR in 0.1MLiPF6/DMSO (from the
CV data from figure 4.6) are depicted in figure 4.7b. At low overpotentials between
about 50 and 150 mV from OCP, the Tafel slope is close to 120 mV/dec. On the
other hand, at high overpotentials, the value is approximately 220 mV/dec. A
120mV/dec Tafel slope is typical of a one-electron process. The subsequent
220mV/dec Tafel slope is due to a second reduction step. The observations in
DMSO/LiPF6 can be summarized by the reactions in Scheme 4.2 involving first the
formation of superoxide O2- first (eq 9) which decomposes (eq 10), or is reduced
further (eq 11), to form O22-. Finally, O2- is formed (eq 12) as the final reduction
product O2.
Scheme 4.2:
Cathodic
O2 + Li+ + e- = LiO2 (Epc1) (9)
2LiO2 = Li2O2 + O2 (chemical) (10)
LiO2 + Li+ + e = Li2O2 (Epc2) (11)
Li2O2 + 2Li+ +2 e- = 2Li2O (Epc3) (12)
Anodic
LiO2 = O2 + Li+ + e- (Epa1) (13)
Li2O2 = O2 + 2Li+ + 2 e- (Epa2) (14)
Li2O = ½O2 + 2Li+ + 2e- (Epa3) (15)
122
Li2O2 as a discharge product of the Li-air battery is well recognized from
Raman spectral analysis of discharged cathodes. Our recent unpublished X-ray
diffraction data for discharged cathodes indicate that Li2O2 and probably Li2O are
discharge products of the Li-air battery. The anodic Tafel slope for Epa1 was
calculated to be 128mV/dec, which is quite similar to Epc1 illustrating the reversibility
of the first one-electron process. The corresponding apparent transfer coefficients (α)
can be calculated from the Tafel slopes. The sum of αc + αa = 1 indicating that the
number of electrons transferred between Epc1 and Epa1 is one. The kinetic
parameters, the cathodic Tafel slope, the cathodic transfer coefficient (αc), the number
of electron transferred (n), and the exchange current density (io) are listed in Table
4.5.
We note here that a reversible reduction of O2 in a Li+-containing electrolyte
is reported here for the first time. The cyclic voltammetric parameters for the
solutions of 0.1M NBu4PF6 and LiPF6 in dimethyl sulfoxide (DMSO), acetonitrile
(MeCN), 1,2 Dimethoxyethane (DME), tetraethylene glycol dimethyl ether
(TEGDME) are summarized in Table 4.3.
Table 4.3: Voltammetric properties of oxygen saturated electrolytes. Scan rate 100mV/s
123
A key difference between O2 reduction in LiPF6-containg DME, MeCN or
TEGDME solution and that in DMSO is the absence of Epc1 and the corresponding
Epa1. Single broad reduction and oxidation peaks are observed in the DME, MeCN or
TEGDME solutions, indicating multiple processes are occurring. We found Epc shifts
toward more negative potentials according to the order
TEGDME<DME<MeCN<DMSO indicating that the reduction of oxygen is hindered
going from DMSO to TEGDME. The reduction of O2 in acetonitrile/LiPF6 is shown
in figure 4.8.
0 1 2 3 4-8e-4
-6e-4
-4e-4
-2e-4
0
2e-4
4e-4
2.50V2.37V2.27V2.10V1.65V0.65V
Voltage (V) Li/Li +
Cur
rent
I(A
mps
/cm
2 )
Epc
Epa3
Epc/2
Epa2
Figure 4.8: Cyclic voltammograms (IR corrected) for the reduction of oxygen in 0.1M LiPF6/MeCN at various potential windows. All scans used a glassy carbon working electrode. Scan rate of 100mV/s.
124
The cathodic peak and half-peak potential are separated (|Epc - Epc/2|) by 220mV
indicating a complex reduction mechanism. Examining the complete CV we notice a
large broad oxidation peak at 3.33V. We studied anodic processes as a function of
cathodic sweep reversal potentials. The CV is first scanned to 2.5V, which is just
after the reduction onset potential. There is little anodic activity at this potential. The
lack of anodic activity indicates that the initial reduction step is irreversible or that the
product undergoes a secondary reaction like that eq.4.10 in scheme 4.2. Increasing
the electrochemical window to 2.37V the half-wave potential (Epc/2) produces an
anodic response Epa2 at 3.25V (grey line), which based on the DMSO data and our
previous results in acetonitrile is believed to be the oxidation of Li2O2. This suggests
that Li2O2 is formed at Epc via the reactions in eq.4.10 and eq.4.11. Anodic peak
capacity increases as the electrode is swept cathodically, closer to the peak potential
of 2.27V (Epc). Maximum anodic activity is reached after sweep reversal at 2.10V a
potential just after Epc. Epa2 begins to broaden and second anodic peak Epa3 emerges
as the potential is scanned cathodically to 1.65V. The presence of this second anodic
peak suggests a third reduction process occurs as the electrode is cathodically
polarized to low potentials, possibly the reduction of Li2O2 to Li2O, eq. 12. Scanning
the electrode to 0.65V, results in disappearance of Epa2 in the following anodic scan.
Oxygen reduction CVs in LiPF6/DME and LiPF6/ TEGDME are illustrated in
figure 4.9a & 9b, respectively. The cathodic peaks are shifted negatively relative to
MeCN, attributed to increase solution resistance and the associated iR polarization.
Little anodic activity is visible prior to arriving at the half wave reduction peak
125
potential Epc/2. The anodic peaks continue to broaden as the CV is scanned towards
Epc.
1 2 3 4-6e-4
-4e-4
-2e-4
0
2e-4
2.4V2.15V2.0V1.82V1.9V
-6e-5
-4e-5
-2e-5
0
2e-5
4e-5
-2.5V-2.2V-2V-1.9V-1.8V-1.5V-1VArgon
Potential (V) vs. Li/Li+
Cu
rrent (Am
ps/Cm
)
Epc
Epa3
Epc/2
Epa2
Epc
Epa3
Epc/2
Epa
Epa2Epa
Epa
B
A
Figure 4.9: Cyclic voltammograms (iR corrected )for the reduction of oxygen in (A) 0.1M LiPF6/DME & (B) 0.1M LiPF6/TEGDME at various potential l windows. All scans used a glassy carbon working electrode. Scan rate of 100mV/s.
126
The broadness of the anodic peak with increasing cathodic potentials indicates that
more than one reduction reaction occurs. The oxidations of these reduction products
occur at Epa1, Epa2 and Epa3. We see that DME & TEGDME differ in that Epa2 in
DME is the predominant peak, while Epa3 manifests itself as the dominant anodic
peak in TEGDME, once the electrode is polarized below Epc. We interpret these
results to mean that the LiO2 formed in the ether electrolytes decompose rapidly to
Li2O2 as we observed in MeCN, and that the Li2O2 is readily reduced to Li2O.
Figure 4.10a shows both a Randles Sevcik plot for a reversible redox couple
(for TBAPF6) and Nicholson plot for an irreversible couple (for LiPF6) in MeCN.
Note the large difference in current for ORR in this electrolyte. A combination of
electrode passivation, oxygen solubility and transfer coefficient contribute to the
decrease of current. The diffusion coefficients of oxygen in both electrolytes are
presented in Table 4.4.
Figure 4.10b shows the scan rate dependence of ORR in both DME and
TEGDME based electrolytes. These plots display an obvious linear relationship
between peak current and scan rate. Both plots clearly obey the Nicholson equation
demonstrating that the oxygen reduction process is totally irreversible in these
electrolytes. This is consistent with the rather small exchange current values derived
below. Cathodic current generated by ORR in the presence of TBA+ is an order of
magnitude larger than the Li+ based electrolyte. Plots of experimental data follow
theoretical n=1 plots quite well although not to the same extent as in DMSO. The
Tafel slopes are much higher as is the case for MeCN (484mv/dec). As the mixed
potential region dominates it is difficult to extract precise kinetic values from these
127
Tafel plots. In such cases it is useful to apply electrochemical impedance
spectroscopy (EIS).
-1.4e-3
-1.2e-3
-1.0e-3
-8.0e-4
-6.0e-4
-4.0e-4
-2.0e-4
0.0
DME Li + ExpDME TBA + ExpTBA+ n=1Li + n=1TEGDME ExpTEGDME n=1
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8-1.8e-3
-1.6e-3
-1.4e-3
-1.2e-3
-1.0e-3
-8.0e-4
-6.0e-4
-4.0e-4
-2.0e-4
0.0
Li + Expn =1n =1TBA+ Exp
Scan Rate (V/s)
Cur
rent
I(A
mps
)
B
A
Figure 4.10: Peak current vs. square root of the scan rate plots for the reduction of oxygen in (A) 0.1 M TBAPF6 & 0.1 M LiPF6/MeCN. n = number of e- (B) 0.1M TBA+ & LiPF6 /DME and 0.1M LiPF6 /TEGDME on GC electrode.
128
Table 4.4: Oxygen Diffusion coefficient in electrolytes
4.3.4 Impedance spectroscopy to determine O2 reduction kinetics
Reaction kinetics can be discerned from faradaic impedance experiments when the
working electrode's potential is held at equilibrium. Departure from equilibrium can
be characterized by the linearized relationship written in terms of the electronic
current as
1s ct
s o
RTR R
C Fiω− = =
(Equation 4.6)
Using this equation the exchange current, and therefore k°, can be evaluated easily (
see equation 7) when the charge transfer resistance Rct is known. Extrapolation of
kinetic data close to equilibrium potential is accomplished by comparing the
calculated data with the experimental results. The data can be analyzed using an
equivalent circuit in which the double layer capacitor is in series with the charge
transfer resistance Rct 21. Plotting Zreal versus ω-1/2 figure 4.11, where Zreal is the real
component of impedance series resistance and ω is frequency. The intercept of this
129
plot is RCt20. The exchange current io is determined from equation 6 and subsequently
the standard rate constant ko is calculated using equation (7).
ooi = nFAk C (Equation 4.7)
0.00 0.02 0.04 0.06 0.08 0.10
ZR
eal / o
hm
0
500
1000
1500
2000
2500
3000
MeCNTEGDMEDME DMSO
Rct= 420 ohm
Rct= 75 ohm
Rct= 2300 ohm
Rct= 308 ohm
1
ω
Figure 4.11: Real impedance versus inverse square root of frequency in 0.1 M LiPF6 DMSO (grey), DME (blue), TEGDME (red) and MeCN (black).
The rate constant provides a true measure of reaction kinetics these values are
tabulated in table 4.5. Table 4.5 shows that the rate constant decreases as the solvents
DN decreases. This dependence implies that the kinetics of the reaction is influenced
strongly on solvent.
130
Table4.5: O2/O2- kinetic parameters of 0.1M Li & TBAPF6
4.3.5 Understanding ORR in non-aqueous electrolytes using Pearson’s
HSAB Theory
The Hard and Soft Acids and Bases (HSAB) theory states Lewis acids and Bases can
classified into hard and soft sub-categories22. Hard acids interact strongly with hard
bases and likewise soft acids interact strongly with soft bases. Hard acids/bases have
a relatively small ionic radius and are difficult to polarize, while soft acids/bases
unusually have larger radii and are easily polarized. Large differences between hard
base solvents and soft-base solvents lead to weaker interactions. Scheme 4.6 shows
order of hardness for both acids and bases.
Scheme 4.3
H+ > Li+ >Fe2+ > Co2+ > Cu2+ > Zn2+ > Ru2+ > Pb2+ > Cu+ > Cd2+ > Au+
Lewis acids
O2-> OH- > F- > Cl- > ClO4- > N2 > NO2 > SO3
2- >Br- > R- > CN- > I- > SCN-
Lewis bases
131
The ions present in the solutions used in this study are the supporting electrolyte ions
TBA+, PF6-, Li+, and the electrochemically generated ions superoxide (O2
-), peroxide
(O22- and monoxide (O2-). The TBA+ is classified as a soft acid due to its large radius
of 0.494 nm (in DMSO)23, and low charge density. It has been shown that
tetraalkylammonium ions, NR4+, are poorly solvated24,25 in organic electrolytes due to
their large size and the small surface charge. A solvent’s basicity is usually
characterized by its donor number (DN) which for the solvents used here follows the
order MeCN(14.1) <TEGDME(16.6) <DME(20.0) <DMSO(29.8). Solvent acidity
can be characterized by its acceptor number (AN) which in these solvents follow the
order DME(10.2) <TEGDME(10.5) <MeCN(18.9) <DMSO(19.3).
In TBA/DMSO electrolytes, although DMSO has a high DN, TBA+ is weakly
solvated. Consequently solvent-TBA+ interactions are weak in the electrolytes
allowing TBA+ to roam more or less as a naked ion26. Among the oxides formed
from the reduction of oxygen, O2- has a relatively large radius and low charge
density; which makes it a moderately soft base. In keeping with the HASB theory,
the naked soft acid TBA+ stabilizes the soft base O2- in the electrolyte with the
formation of an ion pair complex of the type, I
132
Structure I Ion pair between TBA+ and O2- . Nitrogen is blue, carbon is gray and O is red. (Alkyl hydrogens are omitted in the structure)
Reversibility of the O2/O2- redox couple in TBA+ solutions is a result of this stable
solution species I. As O2- is strongly coordinated to TBA+ in I, further reduction of
superoxide to peroxide (O22-) is hindered. The reversibility trend observed in Fig4.1B
appears to follow the acceptor number (AN) trend as the AN increases PF6--solvent
interactions also increase, providing even more TBA+ to interact with O2-. Thus,
DMSO exhibits excellent electrochemical reversibility for the O2/O2- couple. The
lower current in the CV of O2 in DMSO as compared to DME and ACN is probably
due to its lower oxygen solubility. Acetonitrile with high oxygen solubility yields
133
high current for O2 reduction and excellent reversibility in presence of TBA+. In the
case of DME TBAPF6 solutions, both the anodic and cathodic peaks in the CV are
separated by almost one volt, indicative of slow kinetics.
According to the HSAB theory, alkali metal ions are hard Lewis acids and
have a high affinity for hard Lewis bases such as the peroxide and monoxide formed
from the reduction of O2. In electrolyte solutions, the hard Lewis acid Li+ ions are
solvated by the solvents; usually by about four solvent molecules per Li+ to form
solvent separated ion pairs, for example Li+(DMSO)4PF6- in DMSO solutions. The
Li+-solvent bond strength in the complexes would follow the solvent DN scale as
DMSO>MeCN>DME>TEGDME. Nuclear magnetic resonance studies have revealed
that these solvated ion pairs are fluxional complexes even down to – 20 oC. Although
Li+ behaves like a hard acid, its acidity is modulated (or more precisely lowered) by
the strength of the coordination bonds in Li+-(Solvent)n formed with the solvent27.
Since superoxide is a moderately soft base it has low affinity for the hard acid Li+
present in Li+- conducting electrolytes. Consequently, the superoxide formed as the
first reduction product of O2 will want either to decompose or undergo a fast second
reduction to form the hard base, peroxide (O22-), as shown in equations 10 and 11.
Peroxide is a strong Lewis base which wants to be associated with the strong
base Li+. Similarly, the ultimate reduction product of O2, the monoxide O2- is a hard
base with a strong affinity for Li+. Consequently, based on the HSAB theory, the
stable O2 reduction products in Li ion containing electrolyte solutions are Li2O2 and
Li2O.
134
As mentioned above the formation of the Li+-(solvent)n complexes would lower the
acidity of Li+, roughly in proportional to the donor number of the solvent.
In DMSO solutions of LiPF6, the Li+ Lewis acidity is decreased more than in
other solvents due to its higher DN. As a result the superoxide, O2- , formed as the
first O2 reduction product has an increased affinity for these solvated Li+, the O2- is
stabilized longer in solution, in a structure of the type II, reminiscent of the TBA+--O2
complex I.
Structure II Ion pair between solvated Li+ and O2- .(The methyl hydrogen’s are omitted in the structure)
135
This explains the distinct O2/O2- couple seen in the DMSO/LiPF6 solutions. Our
results suggest that depending on the basicity of the solvent measured by its DN, the
superoxide formed as the first reduction product of oxygen will be stabilized to
varying degrees before transforming to O22- via a chemical or an electrochemical
reaction. The multi-step electrochemical reduction of O2 in Li+-containing electrolyte
solutions can be schematically represented in the reaction scheme 4.4.
Scheme 4.4
High DN solvents provide increased stability for complex II because of the modulated
or more precisely decreased Lewis acidity of the hard acid via complex II. In such
electrolytes a distinct O2/O2- reversible couple may be seen in presence of Li+. In
solvents with low DN, the general tendency is for the O2- to quickly decompose or to
undergo fast electrochemical reduction to O22- and further to O2-
136
4.4 Conclusions
Aprotic non-aqueous organic solvents were investigated to determine their influence
on the ORR reactions relevant to the rechargeable Li-air battery. We have
determined how the supporting electrolyte cations, TBA+ and Li+ together with the
solvents comprising the electrolyte solutions influence the nature of reduction
products. In solutions containing TBA+, O2 reduction is a highly reversible one-
electron process involving the O2/O2- couple. On the other hand, in Li+-containing
electrolytes relevant to the Li-air battery, O2 reduction proceeds in a stepwise fashion
to form O2-, O2
2- and O2- as products. These reactions in the presence of Li+ are
kinetically irreversible or quasi-reversible. The stabilization of the one-electron
reduction product, super oxide (O2-) in TBA+ solutions in all of the solvents examined
can be explained using Pearson’s Hard Soft Acid Base (HSAB) theory through the
formation of the TBA+---O2- complex. The HSAB theory coupled with the relative
stabilities of the Li+-(solvent)n complexes existing in the different solvents can also
provide a rational explanation for the different O2 reduction products formed in Li+-
conducting electrolyte solutions. High DN solvents provide increased stability for the
complex [Li+(solvent) n---O2-] because of the modulated Lewis acidity of the hard
acid. In such electrolytes a distinct O2/O2-_ reversible couple may be seen in
presence of Li+. In solvents with low DN, the general tendency is for the O2- to
quickly decompose or to undergo fast electrochemical reduction to O22-. In Li+
electrolytes prepared in low DN solvents O2 may be fully reduced to O2-.
137
4.5 References
(1) Abraham, K. M.; Jiang, Z. J. Electrochem. Soc 1996, 143, 1-5.
(2) Abraham, K. M.; Jiang, Z.; Carroll, B. Chemistry of Materials 1997, 9, 1978-1988.
(3) O 'Laoire, C.; Mukerjee, S.; Abraham, K. M.; Plichta, E. J.;
Hendrickson, M. A. The Journal of Physical Chemistry C 2009, 113, 20127-20134.
(4) Choe, H. S.; Carroll, B. G.; Pasquariello, D. M.; Abraham, K. M.
Chemistry of Materials 1997, 9, 369-379.
(5) Zhang, S. S.; Foster, D.; Read, J. Journal of Power Sources, , 2010, 195, 1235-1240.
(6) Read, J. Journal of The Electrochemical Society 2002, 149, A1190-
A1195. (7) Read, J. J. Electrochem. Soc 2006, 153, A96-A100. (8) Read, J.; Mutolo, K.; Ervin, M.; Behl, W.; Wolfenstine, J.; Driedger,
A.; Foster, D. Journal of The Electrochemical Society 2003, 150, A1351-A1356.
(9) Ye, H.; Xu, J. J. ECS Transactions 2008, 3, 73-81. (10) Kuboki, T.; Okuyama, T.; Ohsaki, T.; Takami, N. Journal of Power
Sources 2005, 146, 766-769. (11) Xu, W.; Xiao, J.; Zhang, J.; Wang, D.; Zhang, J.-G. Journal of The
Electrochemical Society 2009, 156, A773-A779. (12) Kumar, B.; Kumar, J.; Leese, R.; Fellner, J. P.; Rodrigues, S. J.;
Abraham, K. M. Journal of The Electrochemical Society, 157, A50-A54.
(13) Wang, Y.; Zhou, H. Journal of Power Sources, 195, 358-361. (14) Beattie, S. D.; Manolescu, D. M.; Blair, S. L. Journal of The
Electrochemical Society 2009, 156, A44-A47. (15) Débart, A.; Paterson, Allan J.; Bao, J.; Bruce, Peter G. Angewandte
Chemie International Edition 2008, 47, 4521-4524.
138
(16) O'Laoire, C.; Abraham, K. M.; Mukerjee, S. ECS Meeting Abstracts 2009, 804, 404.
(17) Gutmann, V. Coordination Chemistry Reviews 1976, 18, 225-255. (18) Sawyer, D. T.; Chiericato, G.; Angelis, C. T.; Nanni, E. J.; Tsuchiya,
T. Anal. Chem. 1982, 54, 1720-1724. (19) Tsushima, M.; Tokuda, K.; Ohsaka, T. Anal. Chem. 1994, 66, 4551- 4556. (20) Nicholson, R. S.; Shain, I. Anal. Chem. 1964, 36, 706-723. (21) A.J.Bard Electrochemical Methods Fundamentals and Applications; 2
ed.; John Wiley & Sons: New York, 2001; Vol. . (22) Pearson, R. G. Journal of the American Chemical Society 1963, 85,
3533-3539. (23) Paul, R. C.; Johar, S. P.; Banait, J. S.; Narula, S. P. The Journal of
Physical Chemistry 1976, 80, 351-352. (24) Gnanaraj, J. S.; Thompson, R. W.; DiCarlo, J. F.; Abraham, K. M.
Journal of The Electrochemical Society 2007, 154, A185-A191. (25) Tsierkezos, N. G.; Philippopoulos, A. I. Fluid Phase Equilibria 2009,
277, 20-28. (26) Frech, R.; Huang, W. Journal of Solution Chemistry 1994, 23, 469- 481. (27) Abraham, K. M.; Pasquariello, D. M.; Martin, F. J. Journal of The
Electrochemical Society 1986, 133, 661-666. (28) Goldfarb, D. L.; Longinotti, M. P.; Corti, H. R. Journal of Solution
Chemistry 2001, 30, 307-322. (29) Rivas, M. A.; Iglesias, T. P.; Pereira, S. M.; Banerji, N. The Journal of
Chemical Thermodynamics 2006, 38, 245-256. (30) Lago, A.; Rivas, M. A.; Legido, J.; Iglesias, T. P. The Journal of
Chemical Thermodynamics 2009, 41, 257-264. (31) Chemistry of Nonaqueous Solutions:Current Progress; G.Mamantov,
Ed.; Wiley: New York, 1994.
139
(32) Aminabhavi, T. M.; Gopalakrishna, B. Journal of Chemical & Engineering Data 1995, 40, 856-861.
(33) Y.Marcus The Properties of Solvents Wiley, 1998.
140
Chapter 5
A Rechargeable Lithium/TEGDME-LiPF 6/O2 Battery
5.1 Introduction
Rechargeable Lithium-air batteries are attractive electrochemical power
sources because of their potential for ultrahigh energy densities. Despite the
considerable recent research and development interest in these batteries1, the full
energy density promised by the four-electron reduction of O2 to Li2O has not yet been
realized. Furthermore, practically useful electrolytes with low solvent vapor
pressures to enable the operation of Li-air batteries with O2 accessed into the cell
from open air without loosing significant amounts of the solvent in the electrolyte by
evaporation has not yet been satisfactorily demonstrated2-5. Recently, we have
reported the results of our detailed studies of the oxygen reduction reactions (ORR) in
non-aqueous electrolytes showing how the solvent in the electrolyte strongly
influences the reduction products and their rechrgeability6,7. We have shown from
these and earlier investigations8 that polyethylene oxide oligomers are potentially
useful low volatile solvents to build practical Li/air cells. In order to demonstrate this
experimentally and to study the cell chemistry in the absence of catalysts in the
cathode, we have built Li/O2 cells utilizing one of these polyethylene oxide oligomer-
based electrolytes, namely a solution of LiPF6 in tetraethylene glycol dimethyl ether,
CH3O(CH2CH2O)4CH3 (TEGDME) and characterized them. Our principal objectives
of this study have been the following:
141
i) identify the discharge products of the Li/O2 cell in the absence of a catalyst
in the cathode, using a commonly available analytical technique such as X-ray
diffractometry,
ii) determine if the cell is rechargeable without a catalyst in the carbon
cathode, and characterize the relevant cell chemistry, and
iii) elucidate the factors limiting the extended rechargeability of the Li/O2
cell. We have found that the Li/air cell utilizing this electrolyte is rechargeable
though with limited cycle life. We have identified the discharge products of these
Li/air cells from the X-ray diffraction pattern of the discharged carbon cathodes
which is believed to be a first using this technique. We have also made an attempt to
determine the factors affecting the rechargeability of the Li/O2 cells from the AC
impedance spectra of the discharged charged cathode.
5.2 Experimental
5.2.1 Materials
Pursis Tetraethylene glycol dimethyl ether (TEGDME) and anhydrous N-methyl-2-
pyrrolidone (NMP) were purchased from Sigma-Aldrich, Allentown, PA. All
chemicals were dried with Lithium and were stored and prepared in an MBraun dry
box filled with purified argon where the moisture and oxygen content was less than 1
ppm. The dried solvents were stored over 0.3-0.4 nm molecular sieves; and prior to
actual measurements all solvents were degassed under vacuum. Lithium
hexafluorophosphate (LiPF6) (battery grade, >99.9%, H2O< 20ppm) dried under
142
reduced pressure at room temperature was obtained from Novolyte Corporation
Cleveland, Ohio.
5.2.2 Li/O2 Cells
Porous carbon electrodes were prepared as follows. First, ink slurries were prepared
by dissolving a 90 wt% BP2000 carbon black (Cabot Corporation) and 5 wt. % Kynar
PVdF ( Arkema Corporation) in N-methyl-2-pyrrolidone (NMP). Air electrodes were
prepared with a carbon loading of approximately 20.0 mg/cm2 by hand-painting the
inks onto a carbon cloth (PANEX 35, Zoltek Corporation), which was then dried at
180C overnight. The total geometric area of the electrodes was 3.14 cm2.
The Li/O2 test cells were assembled in an argon-filled glove box. The cell
consists of metallic lithium anode and the aforementioned air electrode as a cathode.
A Celgard 2320 separator separated the two electrodes. Both the cathode and the
separator were soaked in a TEGDME/1M LiPF6 solution for a minimum of 24 hours.
An in-house built Li/O2 cell shown in figure 5.1 was used. The cell was placed in an
oxygen filled glove bag where oxygen pressure was maintained at 1atm. Cell
discharge and charge were carried out with an Arbin battery cycler. The AC
impedance was measured on an Autolab PG 30 fitted with a frequency response
analyzer (FRA 2 module) in the range of 0.01 to 106 Hz with an amplitude of 5mV.
Powder X-ray diffraction (XRD) was carried out using a Rigaku RINT 2500X-ray
diffractometer with copper Kα radiation. Scanning electron microscope (SEM)
images and Energy dispersive X-ray spectroscopy were measured using Hitachi SEM
S-4800. All the tests were carried out at room temperature.
143
Figure 5.1: Li-air cell
5.3 Results and Discussion
Li air batteries differ from the conventional batteries in that the air electrode in the
cell continuously reduces oxygen accessed from the environment. Consequently, the
cell is exposed and the loss of solvent from the cell is a concern. Abraham and Jiang9
who demonstrated the first non-aqueous rechargeable Li air cell in 1996 employed a
cell composed of a Li metal anode, a polyacrylonitrile-based gel polymer electrolyte
and a catalyzed carbon cathode. They identified Li2O2 as the main discharge product
of that cell with the aid of Raman spectroscopy. Our recent studies have
demonstrated6,7 that the reduction of O2 can result in Li2O2 and Li2O as stable
products following an initial one-electron product LiO2 which is unstable. The Li/O2
cell’s OCVs calculated on the basis of the reactions yielding these three different
144
products, depicted in equations 1-3, have very similar values indicating that
characterization of discharge product(s) is essential to unequivocally establish the
discharge reaction in a Li-air cell.
Tetra ethylene glycol dimethyl ether (TEGDME) is a polar versatile solvent, which
displays high LiPF6 solubility to produce well-conducting solutions (conductivity
equals, 0.2 mS/cm) with a wide electrochemical window Table 5.1 lists the solvent
properties of TEGDME.
a Rivas et al., Ref.10 b Chemistry of Non aqueous Solutions., Ref.11 c Marcus Properties of Solvents.,Ref.12 d Read,.Ref.13
We first studied the redox electrochemistry of O2 in 1M LiPF6/TEGDME using cyclic
voltammetry on a glassy carbon electrode in order to establish the voltage window of
the electrolyte and to assess the degree of reversibility of the oxygen reduction
reaction. The voltammogram recorded under an atmosphere of argon, shows a nearly
145
5V wide stability window in which there is very little electrochemical activity.
Solvent decomposition is seen as an anodic current at ~4.8V and the onset of lithium
plating is seen only near 0.0V. This seemingly good electrolyte stability window led
us to use this electrolyte in the Li-air battery. Figure 5.2 shows a cyclic
voltammogram (CV) for the reduction of O2 in a 1M LiPF6/TEGDME electrolyte
saturated with oxygen. The peak potential separation ∆Ep between the anodic (Epc =
2.30V) and cathodic (Epa = 4.06V) peaks is almost 1.8V suggesting that O2 reduction
in the presence is only quasi-reversible process.
Figure 5.2: Cyclic voltammograms for the reduction of oxygen in 0.1M LiPF6/TEGDME (Blue) and the argon background (Black). Scan rate 100mV/s.
We recently reported elsewhere on the detailed kinetic analysis of the CV
data6. The chemical reversibility of the observed O2 reduction reaction is an impetus
Potential V ( vs Li/Li+)
0 1 2 3 4 5
Cur
rent
A/c
m2
-3e-4
-2e-4
-1e-4
0
1e-4
2e-4
3e-4
Oxygen Saturated Argon Background
Epc
Epa
146
to construct a Li/O2 cell using this electrolyte and study its discharge reaction and
rechargeability, and characterize the discharge products formed on the carbon
cathode.
5.3.1 Li/O2 Cell Discharge and Charge Behavior.
The full discharge curves for two Li/O2 cells obtained with BP 2000 carbon
electrodes exposed to a dry oxygen atmosphere are depicted in figure 5.3. The air
electrode had loadings between 7-8.5 x 10-4g/cm2, respectively, of the BP 2000
carbon on Panex carbon cloth.
Figure 5.3 : Li/O2 cell discharge curves at 0.25 (blue) & 0.16 (black) mA/cm2 in 1M LiPF6/TEGDME. Capacities are expressed per gram of BP 2000 carbon in the electrode.
The open-circuit voltages (OCV) of the cells varied slightly, 3.37 V (black)
and 3.18V (blue) depending on cell setup, oxygen saturation and atmospheric
Discharge Capacity (mAh/g)
0 500 1000 1500 2000 2500 3000
Vol
tage
(V
)
1.0
1.5
2.0
2.5
3.0
3.5
0.16mA/cm2 0.25mA/cm2
147
variables. The working voltages of the cells differed on average by 150mV similar to
the difference in the OCVs. The discharge current density affects the specific
capacity of the Li-air cell. The discharge capacity at 0.16mA/cm2 is 2,760mAh/g of
the BP 2000 carbon in the cathode. Increasing the current to 0.25mA/cm2 decreases
the specific capacity to 1452mAh/g of the carbon. These values are very high
considering that there is no catalyst in the cathode. It should be noted some
discharge of O2 occurs also on the carbon cloth that is used as the current collector
but the capacity was small 75-90mAh/gram compared to that on the high surface area
carbon. As a result, we have expressed observed the discharge capacity only in terms
of capacity per gram of the high surface BP 2000 carbon.
Following full discharge at 0.25 mA/cm2, the cell was disassembled in the
glove box and the carbon cathode washed in acetonitrile to remove LiPF6. Figure 5.4
shows the x-ray diffraction pattern of the fully discharged carbon electrode. X-ray
diffraction was recorded using Cu Kα radiation with normal (θ - 2θ) scanning at very
slow speed, essential to obtain an assignable pattern. The spectrum can be assigned
to that of Li2O2 based on the JCPDS data card (# 7400115). We conclude that the
discharge plateau observed at 2.5V is the result of Li2O2 formation confirming the
earlier observation made using Raman spectroscopy.
148
Figure 5.4: XRD pattern of fully discharged O2 cathode in 1M LiPF6/TEGDME. The lines represent the JCPDF pattern for Li2O2.
Figure 5.5 depicts the full discharge of another cell discharged at 0.25 mA/cm2. Its
discharge capacity is 1452mAh/g and a subsequent charge to 4.8 V yields charge
capacity of 408mAh/g. In the following discharge a capacity of 488mAh/g is
obtained from this cell. The second charge step was shorter with only 200mAh/g
capacity before the voltage runs to 4.9V and the cell ultimately fails.
149
Figure 5.5: Full Discharge of Li/O2 cell discharge at 0.16 mA/cm2 in 1M LiPF6/TEGDME. Following discharge the cell was charged to 4.5V
Several cells were cycled at various depths of discharge (DOD), and charge,
to follow the impedance of the cell as a function of charge and discharge to ascertain
the causes of poor cell rechargeability. These shallow DOD cycles also allowed us to
investigate the impact of current density on rechargeability, and to follow cell
impedance as a function of cycle life as cycles could be accumulated quickly to
provide information on factors affecting cell deterioration.
0 5 10 15 20 251
2
3
4
5
6
Time (h)
Vol
tage
150
0 10 20 30 40 500
20
40
60
80
100
120
140
160
180
200
DischargeCharge
Cycle Number
Cap
acity
(mA
h/g)
Discharge/Charge 0.13mA/cm2
B
Capacity (mAh/g)
0 50 100 150 200
Vol
tage
(V
)
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Cycle 1Cycle 10Cycle 13Cycle 14Cycle 15Cycle 20Cycle 30Cycle 41
A
Figure 5.6: A) The cycling data for a 1M LiPF6/TEGDME electrolyte oxygen cell at room
temperature. The cell was discharged and charged for 2 hours at 0.13 mA/cm2. Capacities are expressed per gram of BP 2000 carbon + PVDF in the electrode. B) Discharge/Charge capacities as a function of cycle number for the same cell.
151
0 200 400 600 800 1000
2.5
3.0
3.5
4.0
4.5
Cycle 1Cycle 2Cycle 3Cycle 4Cycle 5Cycle 6Cycle 7
Capacity mAh/g
Vol
tage
(V
)
0 2 4 6 8 100
200
400
600
800
1000
DischargeCharge
Cycle Number
Cap
acity
mA
h/g
B
A
Figure 5.7: A) The cycling data for a 1M LiPF6/TEGDME electrolyte oxygen cell at room
temperature. The cell was discharged and charged for 14 hours at 0.13 mA/cm2. Capacities are expressed per gram of BP 2000 carbon + PVDF in the electrode. B) Discharge/Charge capacities as a function of cycle number for the same cell.
Figure 5.6a illustrates capacity versus cycle number for a cell discharged and charged
for 2-hour periods at 0.13mA/cm2. The voltage is monitored as a function of time.
152
The average discharge potential varied between 2.7V for the initial cycles and
dropped to 2.45V at the later stages of the cell’s life. The charge plateau steadily
increased from 3.2 V initially to 4.7V before cell termination. Figure 5.6b presents
the discharge/charge capacity as a function of cycle life. Cycling is demonstrated
over 40 cycles during which 100% columbic efficiency is achieved with a capacity
utilization of 175mAh/g. The discharge fell below two hours at cycle 41.
High discharge/discharge capacity was maintained after increasing discharge
time to 14 hours but the cell exhibited decreasing capacity with increased charge
voltage polarization after about 4 cycles (Fig5.7a-b). The discharge profile remains
unchanged after the first cycle until sudden cell failure observed during cycle 7,
where capacity drops to 400mAh/g. Efficient cycling was observed during the first
three charge cycles; however charge profile polarization increases with increasing
cycle number further. As polarization increases capacity drops and charge voltage
increases. The sudden drop in capacity at the 8th discharge coupled with the
impedance data and the physical examination of the Li anode after cycling ( see
below) led us to believe Li anode failure as a primary contributor to cell failure.
Figure 5.8 illustrates discharge versus capacity utilization of several cells
cycled at various current densities. By lowering the current density from 0.25 to
0.13mA/cm2 the charge efficiency of the cell improves to 100% rechargeability.
Clearly, rechargeability is affected by the current density. At the highest
charge/discharge current density of 0.25 mA/cm2 the cell’s discharge capacity
remains steady for the first 6 cycles and drops off sharply thereafter. The recharge on
other hand is less than 100 % starting with the first cycle with precipitous loss
153
occurring in the rest of the cycles. The poor recharge efficiency at the highest current
density may be attributed primarily to the inability to oxidize the non-conducting
discharge product Li2O2 deposited in the pores of the carbon cathode due perhaps to
the increased resistance and the associated over-voltage of the electrode. However,
the impedance data discussed below as well as post-test examination of the Li anode
revealed that the deterioration of the Li anode is also a major contributing factor for
the capacity decline and eventual failure of the Li/air cell
Figure 5.8: Discharge curves of the lithium air cell at various current densities in 1M LiPF6/TEGDME oxygen cell at room temperature. Capacities are expressed per gram of
carbon in the electrode. (Red) 0.25mA/cm2, (Blue) 0.13mA/cm2, (Black) 0.07mA/cm2.
0 2 4 6 8 10 120
100
200
300
400
500
600
Cycle Number
Cap
acity
(mA
h/g)
Discharge/Charge 0.25mA/cm2
Discharge/Charge 0.07mA/cm2
Discharge/Charge 0.13mA/cm2
154
5.3.2 Factors affecting the Cycle Life of the Li/O2 Cell
Figure 5.9a shows the complete discharge profile of a Li/O2 cell to 0.85V. As
expected a constant discharge plateau is observed above 2.5V. After 1500mAh/g the
cell voltage drops gradually. A second voltage region emerges below about. 2 V.
Solvent decomposition is not the reason for the small distinct voltage region below 2
V. fig 5.9). We know from our previous work that Li2O2 is reduced to Li2O at
potentials below or close to 2V (reaction 3). This decomposition process was
discussed in previous chapters. Fig 5.12b shows that there is a significant increase in
the cell impedance during the full discharge (Rp= 250Ω). This confirms the poor
rechargeability of the system after full discharge.
155
0 50 100 150 200 250 3000
20
40
60
80
100
FreshPost Discharge
Z' / ohm
-Z''
/ ohm
Discharge Capacity (mAh/g)
0 500 1000 1500 2000 2500 3000 3500
Vol
tage
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
B
A
Figure 5.9: (a) Full discharge of Li/air cell in 1M LiPF6/TEGDME (-0.13mA/cm2). (b) Nyquist plot
156
Figure 5.10: Nyquist impedance plots of the Li-air battery cycled at 2h discharge at the ends of various discharges (9a) and charge (9b). Also for the cell cycled at 14h discharge depths (9c) at the end of different discharges. . The data were fitted by using a RC equivalent-circuit model.
Figure 5.10 displays typical AC impedance spectra recorded at various stages in the
cycle life of the Li air cells displayed in figures 5.6 & 5.7. Generally, the spectra
displayed show an offset semi-circle at high frequencies. At low frequencies the
semi-circle connects with a line inclined at approximately 45o to the x-axis. Initially,
the impedance as indicated by the diameter of the semi-circle increases with cycle
number (fig 5.10). However, after cycle 4 in the cell in fig 5.10a and after cycle 3 for
the cell described the fig 10b impedance drops slowly. Several qualitative changes in
the spectra recorded intermittently during the course of the cell discharge are
157
noticeable. First there is a large increase in the diameter of the semicircle without any
significant change in the position of its first intersection with the x-axis. The
diameter of the semi-circle reached a maximum at about the fourth (fig 5.10a) and
sixth (fig 5.10b) cycles. Contrary to what we believed prior to the experiment there
was no significant difference in the impedance between a discharge and the following
charge.
Fitting the impedance spectrum to an appropriate equivalent circuit is difficult
since the cell does not have a reference electrode which prohibits assigning the
contributions of the anode and cathode interfacial reaction impedances to the total cell
impedance. A simple equivalent circuit shown in fig 5.10d may describe the
observed impedance spectrum at the early stages of the cycling. The circuit consists
of an Ohmic resistance (Rs) due to electronic resistances of the electrodes and their
contacts to the current collectors, and electrolyte resistance, that is in series with a
constant phase element that represents the capacitive contributions of the two
electrodes in parallel with the polarization (charge transfer) resistances (Rp) at the
bulk electrodes. Clearly the charge transfer resistances of the reactions at both the
anode and cathode contribute to this. Without a reference electrode we can only
discuss overall cell polarization impedance and with that in mind the capacitance is
represented by a constant phase element Cp. The pre-cycling spectrum of the cell
reveals one semicircle with the two intersection Rs = 11.50 and Rp= 14.00 Ohm. The
linear Warburg element following the semicircle may be attributed to the diffusion of
electroactive species to the electrode. The polarization resistance in the 2 hour cycled
cell increased after the first discharge to Rp= 46.4 ohm. Overall impedance of the air
158
electrode gradually increases with cycle number. However towards the end of the
cycle life of the cell the initial semi circle becomes depressed and a second semi-
circle emerges. There is little change in the intercept Rs at high frequencies. The
impedance increase at the end of discharge is attributed to the deposition of the
discharge products in the pores and the surface of the carbon electrode, resulting in
sluggish ORR kinetics and diffusion of electroactive species to the electrode surface.
As these spectra are of the whole cell, the reactions at the Li electrode, the changing
morphology of the plated lithium and the surface films formed on it also contribute to
this polarization.
Figure 5.11 displays the photographs of the Li anode before and after cycling.
The figure displays the totally changed morphology of the plated Li and shows the
granular lithium particles on the surface of the metal after many charge/discharge
cycles. The growth results in an increase in overall surface area of the Li anode. This
in turn leads to a decrease in the resistance of the surface films on the Li anode and an
overall lowering of the anode’s contribution to the total impedance. This would
explain the drop in the total cell impedance after several cycles. The evidence
suggests that the point where the total cell impedance begins to decrease is a marker
of a significant change in the morphology of the Li anode
159
A B
Figure 5.11: (a) Fresh Lithium. (b)Lithium anode after cycling
SEM micrographs of surface morphology of the O2 cathode electrode are shown in
figure 5.12 (a,b). Figure 5.12a shows individual particles of BP2000 carbon on the
Panex substrate on the undischarged electrode. Average particle size of BP2000 is
12nm. Figure 5.12b reveals a much different surface after discharging the electrode
at a discharge rate at 0.13 mA/cm2 in an oxygen atmosphere. It is clear that the
discharge products are evenly deposited on both BP2000 and the Panex substrate
resulting in high specific capacity. The deposit as determined (Fig.5.11d) by energy-
dispersive X-ray spectroscopy (EDAX) analysis is extremely oxygen rich, which
supports the presence of Li2O2 by XRD analysis. The EDAX also reveals the
presence of traces of LiPF6 in the electrode.
160
A B
B DD
C
Figure 5.12: SEM micrographs of the O2 cathode (11a) fresh (11b) discharged. Scale bar is 1 μm. Energy-dispersive X-ray spectroscopy (EDAX) (11c) fresh (11d) discharged at 0.13 mA/cm2 in oxygen.
5.4 Conclusions
The use of the low volatile electrolyte TEGDME-LiPF6 allowed us to study
the discharge reaction and rechargeability of the Li/O2 cell substantially without the
uncertainties associated with solvent evaporation on cell failure. The cell was
fabricated sans electrocatalytic catalyst in the carbon cathode in order to characterize
cell chemistry in the baseline Li/O2 cell. From the X-ray diffraction patterns of
discharged carbon electrodes we identified Li2O2 in cells discharged to 2.0 V, and,
additionally, Li2O in cells discharged to 1.0 V. The rechargeability of the
uncatalyzed cell is limited which to a large extent is attributed to the poor cycling
161
efficiency of the Li anode in addition to the impedance associated with the Li2O2
deposit in the carbon cathode
5.5 References
(1) Armand, M.; Tarascon, J. M. Nature 2008, 451, 652-657. (2) Xu, W.; Xiao, J.; Wang, D.; Zhang, J.; Zhang, J.-G. Journal of The Electrochemical Society 2010, 157, A219-A224. (3) Zhang, S. S.; Foster, D.; Read, J. Journal of Power Sources, 2010 195, 1235-1240. (4) Zhang, D.; Li, R.; Huang, T.; Yu, A. Journal of Power Sources, 2010-195, 1202-1206. (5) Zhang, T.; Imanishi, N.; Shimonishi, Y.; Hirano, A.; Takeda, Y.; Yamamoto, O.; Sammes, N. Chemical Communications, 2010 46, 1661- 1663. (6) O 'Laoire, C.; Mukerjee, S.; Abraham, K. M.; Plichta, E. J.; Hendrickson, M. A., The Journal of Physical Chemistry C, 2009, 113, 20127-20134. (7) O'Laoire, C.; Abraham, K. M.; Mukerjee, S. ECS Meeting Abstracts 2009, 804, 404. (8) Abraham, K. M.; Jiang, Z.; Carroll, B. , Chemistry of Materials 1997, 9, 1978-1988. (9) Abraham, K. M.; Jiang, Z. J. Electrochem. Soc .1996, 143, 1-5. (10) Rivas, M. A.; Iglesias, T. P.; Pereira, S. M.; Banerji, N. ,The Journal of Chemical Thermodynamics ,2006, 38, 245-256. (11) Chemistry of Nonaqueous Solutions:Current Progress; G.Mamantov, Ed.; Wiley: New York, 1994. (12) Y.Marcus The Properties of Solvents Wiley, 1998.
162
(13) Read, J.; Mutolo, K.; Ervin, M.; Behl, W.; Wolfenstine, J.; Driedger, A.; Foster, D., Journal of The Electrochemical Society 2003, 150, A1351-A1356.
163
Chapter 6
Summary and Future Directions
6.1 Summary
High energy density Li- Air batteries once a laboratory curiosity, are now the
focus of serious high-level research. The prospect of cheap high-density energy
conversion and storage is irresistible to such industrial giants as Toyota and IBM,
both of whom have made a considerable investment in the technology. In the space
of three years the number of peer-reviewed articles on the subject has tripled. This
thesis endeavors to build upon all previous studies and contribute to furthering the
science underlying this battery. Probing electrochemical interfaces using modern
electrochemical techniques along with conventional characterization methods yielded
a wealth of information regarding the chemical and electrochemical processes in the
battery. A highlight of the work is the development of a fundamental mechanistic
scheme for oxygen reduction reactions in non-aqueous electrolytes and a
methodology for the systematic design of an optimal Li-Air battery electrolyte
(chapter 5). In this chapter I summarize the findings and results acquired throughout
my research.
6.2 Salt Effects on ORR
We found that the reduction and subsequent oxidation of O2 in acetonitrile-
based electrolytes is strongly influenced by the cation of the conducting salt used.
Oxygen reduction reactions in Li salt solutions result in irreversible or quasi-
164
reversible electrochemistry and passivation of the electrodes by the reduction
products. On the other hand, ORR in TBA salt solutions exhibits a highly reversible
oxygen redox couple. A practical outcome of this observation is that it would be
advantageous to use a mixture of Li and K and/or TBA salts as supporting
electrolytes in order to dissolve the oxygen reduction products in a Li oxygen battery,
at least when the battery is used a primary battery. Increasing the solubility of the
reduction products would delay the passivation of the porous electrode that in turn
would increase the amount of oxygen that can be reduced to deliver higher capacity.
Dissolving the reduction products could also promote reversibility of O2 reduction,
which would increase the battery’s rechargeability. Our results also show that vital
electrochemical kinetic data provide a platform to quantify catalytic effects on the
ORR reaction. Kinetic data are relevant to the studies of the Li-air battery as well a
others containing soluble electrode materials, especially for battery simulation studies
aimed at understanding the performance of practical batteries, and generally for the
development of improved materials.
6.3 Solvent Effects on ORR
Chapter 4 examines the electrochemical reduction of oxygen in various
TBAPF6 & LiPF6-based organic electrolytes in a series of solvents selected on the
basis of their widely varied Donor Numbers.. Stable TBA+---O2- complexes are
formed in all organic TBA+ electrolytes solutions. This is explained by Pearson’s
Hard Soft Acid Base (HSAB) theory. The HSAB theory also provides a rational
explanation for the influence of both conducting salts and the organic solvents on the
nature of the reduction products. High DN solvents provide increased stability for the
165
complex [Li+(solvent)n---O2-] because of the modulated decreased Lewis acidity of
the hard acid Li+ In such electrolytes a distinct O2/O2 reversible couple may be seen
in presence of Li+. In solvents with low DN, the general tendency is for the O2- to
quickly decompose or to undergo fast electrochemical reduction to O22-. In Li+
electrolytes prepared in low DN solvents, O2 may be fully reduced to O2-.
6.4 Experimental Li-Air Cells
The use of the low volatile electrolyte TEGDME-LiPF6 allowed us to study
the discharge reaction and rechargeability of the Li/O2 cell substantially without the
uncertainties associated with solvent evaporation on cell failure. The cell was
fabricated with out an electrocatalytic catalyst in the carbon cathode in order to
characterize cell chemistry in the baseline Li/O2 cell. From the X-ray diffraction
patterns of discharged carbon electrodes we identified Li2O2 in cells discharged to 2.0
V, and, additionally, Li2O in cells discharged to 1.0 V. The rechargeability of the
uncatalyzed cell is limited which to a large extent is attributed to the poor cycling
efficiency of the Li anode in addition to the impedance associated with the Li2O2
deposit in the carbon cathode. Our continuing effort to obtain a clear understanding
of the roles of conducting salt and the solvent is expected to result in the
identification of an optimum electrolyte solution for the non-aqueous Li-air battery
6.5 Future Directions for Li-Air Research
We have only begun to scratch the surface of the fledging Li-Air battery
research field. Realization of the Li-Air dream will require a long-term research
166
initiative. The development of a practical rechargeable Li-air battery will require
active research in the fields of catalysis development and electrolyte stability
especially towards lithium anode. The prospect for design of electrocatalysts
specifically for Li2O2 and Li2O oxidation is challenging. Viable candidates for this
work are the macro cycle complexes such as porphyrins, bimetallic porphyrins and
phthalocyanines. Nano-porous amorphous manganese oxide is emerging as a
promising electrocatalyts in non-aqueous electrolytes. Electrocatalytic activity
towards oxide oxidation is imperative to maximizing efficiency. Our electrochemical
studies have laid the groundwork for such electrocatalytic studies. Stable electrolytes
are crucial to the success of Li-air. As demonstrated in chapter 5 low volatile
solvents such TEGDME are possible candidates. However superior solvents blends
should be considered which combine beneficial characteristics of multiple solvents
such as low volatility, high-oxygen solubility and low viscosity. Possible candidates
are room temperature ionic liquids (RTIL). These solvents can be designed to
incorporate multiple desirable characteristics of a Li-air electrolyte. Ultimately,
packaging of practical cells and batteries would require devoting considerable
resources to engineering development.
167
Biographical Information
Name: Cormac Ó Laoire
Birthplace: Cork, Ireland
Education
Ph.D., Physical Chemistry, May 2010
Northeastern University, Boston, MA
(INVESTIGATIONS OF OXYGEN REDUCTION
REACTIONS IN NON-AQUEOUS ELECTROLYTES AND
THE LITHIUM-AIR BATTERY)
M.Sc., Materials Science, May 2004
University College Cork, Cork, Ireland
(Thesis: Analysis of Acid Passivation of Stainless Steel)
B.Sc., Chemistry, May 2001
University College Cork, Cork, Ireland
168
Experience:
Sept 2006-Present
Lithium air batteries are being developed under the direction of
Professors K.M. Abraham and Sanjeev Mukerjee. My research
focused on elucidating the kinetics and mechanism of the oxygen
reduction reaction in the non-aqueous environment with
particular emphasis on the roles of ion conducting salts and the
non-aqueous solvents. Electrochemical and kinetic
characterizations are performed with a wide array of
electrochemical and analytical techniques including cyclic
voltammetry, rotating disk electrode voltammetry,
chronocoulometry and charge/discharge cycling of Li-air cells.
Sept 2004-Sept 2006
Novel non-noble metal chalcogenide clusters development for
oxygen reduction reaction in fuel cell applications under the
direction of Professor Sanjeev Mukerjee. Also during this time
period I investigated carbon-based materials for Li-ion batteries.
Structural characterizations were accomplished via in situ
EXAFS, XANES and XRD at the National Synchrotron Light
Source (Brookhaven National Labs, Upton, NY).
169
Appointments:
2006-present Research Assistant
Northeastern University, Boston, MA
2004-2006 Teaching Assistant- General and Physical Chemistry
Northeastern University, Boston, MA
2002-2004 Research Assistant-
University College Cork, Cork, Ireland
Awards: 2009 ECS Battery Division Travel Grant Award (Vienna 2009)
2010 Northeastern University Dissertation completion fellowship
Publications:
“Influence of Nonaqueous Solvents on the Electrochemistry of Oxygen in
the Rechargeable Lithium-Air Battery”
O’Laoire, C.; Plichta, E.; Hendrickson, M.; Mukerjee, S.; Abraham, K. M.
(Accepted) J. Phys. Chem. C April 2010
“Elucidating the Mechanism of Oxygen Reduction for Lithium-Air
Battery Applications” Cormac O. Laoire, Sanjeev Mukerjee, K. M.
Abraham, Edward J. Plichta, Mary A. Hendrickson J. Phys. Chem. C 2009
113 (46), 20127-20134
170
“Electrochemical studies of ferrocene in a lithium ion conducting organic
carbonate electrolyte” O’Laoire, C.; Plichta, E.; Hendrickson, M.;
Mukerjee, S.; Abraham, K. M. Electrochimica Acta 2009, 54, 6560-6564.
“Electrochemical kinetics and X-ray absorption spectroscopy
investigations of select chalcogenide electrocatalysts for oxygen reduction
reaction applications” Ziegelbauer, J. M.; Murthi, V. S.; O'Laoire, C.;
Gullá, A. F.; Mukerjee, S. Electrochimica Acta 2008, 53, 5587-5596.
“Chalcogenide electrocatalysts for oxygen-depolarized aqueous
hydrochloric acid electrolysis”----Ziegelbauer, J. M.; Gullá, A. F.;
O'Laoire, C.; Urgeghe, C.; Allen, R. J.; Mukerjee, S. Electrochimica Acta
2007, 52, 6282-6294.
“Analysis of Acid passivation of Stainless Steel”----O'Laoire, C.;
Timmins, B.; Kremer, L.; Holmes, J. D.; Morris, M. A. Analytical Letters
2006, 39, 2255 - 2271.
Skills:
Synthetic
Solid State synthesis of carbon materials for Li-ion batteries.
171
Electrochemical Instrumentation
Battery Cyclers (Arbin), Potentiostat/Galvanostats (BAS,
Autolab, PAR), Rotating Disk Electrodes (Autolab and BAS).
Other Instrumentation
SEM, IR, ESR, NMR, XRF Extended X-Ray Fine Absorption
(EXAFS) and High Energy X-ray diffraction.
Computer
Instrument programming skills include C++, and Visual Basic in
addition to the following molecular modeling programs: Amber,
DLPoly, and BioMedCAChe.
Presentations:
‘Solvent and Conducting Salt Effects on the Oxygen
Reduction Mechanism in the Non-Aqueous Lithium
Air Battery’, Cormac O Laoire, K.M. Abraham and Sanjeev
Mukerjee, Rechargeable Lithium Ion Batteries, New Battery
Materials and Systems, 216th Meeting of the Electrochemical
Society, Vienna, Austria, 2009
172
‘The effect of solvent on the Oxygen Reduction
Mechanism in the Non-Aqueous Lithium Air Battery’, Cormac O
Laoire, K.M. Abraham and Sanjeev Mukerjee, General Battery
Session, Non-Aqueous Systems, 214th Meeting of the
Electrochemical Society, Honolulu, HI, 2008
‘Oxygen Reduction Mechanism in the Non-Aqueous
Lithium Air Battery’, Cormac O Laoire, K.M. Abraham and
Sanjeev Mukerjee, General Battery Session, Supercapacitors,
hybrids, and batteries 213th Meeting of the Electrochemical
Society, Phoenix, AZ, 2008
